1(self.webpackChunk_N_E=self.webpackChunk_N_E||[]).push([[5944],{54580:function(e,a,t){(window.__NEXT_P=window.__NEXT_P||[]).push(["/data-service/web-data-for-ai",function(){return t(87274)}])},91563:function(e,a,t){"use strict";var i=t(11527),s=t(75273),n=t.n(s);let r=e=>{let{items:a,containerClassName:t,ariaLabel:s="Service metrics"}=e,r=a.length>4?n().gridThree:n().gridFour;return(0,i.jsx)("div",{className:n().bandWrap,children:(0,i.jsx)("div",{className:t,children:(0,i.jsx)("div",{className:n().bandSurface,"aria-label":s,children:(0,i.jsx)("div",{className:"".concat(n().grid," ").concat(r),children:a.map((e,a)=>{let t=e.key||[e.prefix,e.value,e.suffix,e.description,String(a)].filter(Boolean).join("-");return(0,i.jsxs)("article",{className:n().card,children:[(0,i.jsxs)("div",{className:n().measure,children:[e.prefix?(0,i.jsx)("span",{className:n().prefix,children:e.prefix}):null,(0,i.jsx)("span",{className:n().value,children:e.value}),e.suffix?(0,i.jsx)("span",{className:n().suffix,children:e.suffix}):null]}),(0,i.jsx)("div",{className:n().description,children:e.description})]},t)})})})})})};a.Z=r},86744:function(e,a,t){"use strict";var i=t(11527),s=t(95511),n=t(49865),r=t(16040),o=t(82187),c=t.n(o),d=t(8857),l=t(94826),u=t.n(l);let p=e=>{let{faqList:a,title:t,compact:o=!1,children:l}=e;return(0,i.jsx)("div",{className:c()(u().faq,{[u().compact]:o}),children:(0,i.jsxs)(s.W,{size:o?void 0:"sm",className:u().faq_wrapper,children:[(0,i.jsx)(n.D,{order:2,className:u().title,children:t}),l,(0,i.jsx)(r.U,{variant:"separated",className:u().accordion,children:a.map(e=>(0,i.jsxs)(r.U.Item,{className:u().item,value:e.value,children:[(0,i.jsx)(r.U.Control,{className:u().item_question,children:e.question}),(0,i.jsx)(r.U.Panel,{children:(0,i.jsx)("div",{className:u().answer,children:(0,d.ZP)(e.answer)})})]},e.value))})]})})};a.Z=p},87274:function(e,a,t){"use strict";t.r(a),t.d(a,{__N_SSP:function(){return D}});var i=t(11527),s=t(8968),n=t(46240),r=t(91563),o=t(38324),c=t(8824),d=t(86744),l=t(89501),u=t(57402),p=t(67422),m=t.n(p),f=t(83487),h=t.n(f);let g={"@context":"https://schema.org","@type":"Service",name:"Custom Web Datasets for AI Training",alternateName:["Custom Web Datasets for AI Training","AI Training Data Service","LLM Fine-tuning Dataset Provider","RAG Data Feed Service","Domain-Specific Web Data for Machine Learning","Custom Dataset for AI Training","AI Agent Data Feed Service","Structured Web Data for LLM","Common Crawl Alternative for Domain Data","Managed Web Data for AI Startups","JSONL Dataset from Web Data Service","Pretraining Corpus Collection Service","AI training dataset service","LLM fine-tuning data provider","RAG knowledge base data pipeline","web scraping for machine learning","domain-specific corpus collection service","structured web data for GPT fine-tuning"],serviceType:"Managed custom web datasets for AI training, RAG, LLM fine-tuning, and AI agent data pipelines",provider:{"@type":"Organization",name:"Octoparse",url:"https://www.octoparse.com",contactPoint:{"@type":"ContactPoint",contactType:"Sales",email:"[email protected]",availableLanguage:["English","Chinese"]}},areaServed:[{"@type":"Country",name:"United States"},{"@type":"Country",name:"China"},{"@type":"Country",name:"Singapore"},{"@type":"Country",name:"United Kingdom"},{"@type":"Country",name:"Australia"},{"@type":"Country",name:"Germany"},{"@type":"Place",name:"Global"}],url:"https://www.octoparse.com/data-service/web-data-for-ai",description:"Octoparse builds custom web datasets for AI training, RAG knowledge bases, LLM fine-tuning, AI agents, and domain-specific model evaluation. Teams define their target domains, sources, fields, freshness requirements, and delivery format; Octoparse handles collection, deduplication, schema normalization, provenance tagging, QA, and delivery in JSONL, Parquet, CSV, JSON, S3, GCS, Snowflake, BigQuery, Redshift, API, or webhook. Unlike generic crawls such as Common Crawl, Octoparse delivers purpose-built web datasets with source URL, publish timestamp, language tag, domain classification, and token count estimates on a one-time or recurring schedule. Projects start from $699/project; free scoped samples are delivered within 1-2 business days.",ke
1ywords:["custom web datasets for AI training","LLM fine-tuning dataset provider","RAG knowledge base data pipeline","JSONL dataset from web data service","domain-specific corpus collection service","AI agent data feeds","web data feed to Snowflake","web data feed to BigQuery","Common Crawl alternative for domain data"],offers:{"@type":"Offer",priceCurrency:"USD",price:"699",description:"Web data for AI workflows â from $699/project for one-time datasets, $599/month for recurring feeds. Enterprise custom pricing available.",availability:"https://schema.org/InStock"}},w={"@context":"https://schema.org","@type":"ItemList","@id":"https://www.octoparse.com/data-service/web-data-for-ai#public-ai-dataset-proofs",name:"Public sample datasets connected to Octoparse custom web datasets for AI training",description:"Public-safe sample datasets that show structured fields, provenance context, workflow metadata, and delivery patterns for AI-ready web data pipelines.",itemListElement:[{"@type":"ListItem",position:1,item:{"@type":"Dataset",name:"E-commerce Visual Matching Dataset",url:"https://www.kaggle.com/datasets/octoparsedataservice/ecommerce-visual-matching-dataset",isBasedOn:"https://www.octoparse.com/data-service/ai-visual-product-matching",variableMeasured:["source_product_id","candidate_product_id","similarity_score","match_status","reject_reason"]}},{"@type":"ListItem",position:2,item:{"@type":"Dataset",name:"TikTok Brand Monitoring Beauty Sample",url:"https://www.kaggle.com/datasets/octoparsedataservice/tiktok-brand-monitoring-beauty-sample",isBasedOn:"https://www.octoparse.com/data-service/social-media-monitoring",variableMeasured:["video_url","creator","brand","caption","engagement_signal","captured_at"]}},{"@type":"ListItem",position:3,item:{"@type":"Dataset",name:"Temu E-commerce Pricing Workflow Sample",url:"https://www.kaggle.com/datasets/octoparsedataservice/temu-ecommerce-pricing-workflow-sample",isBasedOn:"https://www.octoparse.com/data-service/competitor-price-monitoring",variableMeasured:["product_url","sku","price","seller","stock_status","captured_at"]}}]},b={"@context":"https://schema.org","@type":"FAQPage",mainEntity:[{"@type":"Question",name:"What are custom web datasets for AI training?",acceptedAnswer:{"@type":"Answer",text:"Custom web datasets for AI training are scoped, structured datasets collected from approved public web sources for model training, LLM fine-tuning, RAG knowledge bases, AI agent monitoring, evaluation sets, or domain-specific model adaptation. Octoparse defines the sources, fields, schema, freshness cadence, provenance metadata, and delivery format with your team before collection begins."}},{"@type":"Question",name:"How is this different from Common Crawl?",acceptedAnswer:{"@type":"Answer",text:"Common Crawl is a broad monthly web crawl covering the whole internet â useful as a base for pretraining but not domain-specific, and requiring a major pipeline to extract and clean the relevant fraction. Octoparse delivers purpose-built datasets scoped exactly to your target domain, sources, and fields: structured, deduplicated, normalized, and delivered in JSONL or Parquet â not raw HTML requiring another engineering project to make usable."}},{"@type":"Question",name:"What output formats are supported for AI training data?",acceptedAnswer:{"@type":"Answer",text:"Octoparse delivers structured web data as JSONL (standard for LLM fine-tuning), Parquet (for large-scale data processing), CSV, JSON, or direct push to cloud storage (S3, GCS) and data warehouses (Snowflake, BigQuery, Redshift). Every delivery includes provenance metadata: source URL, timestamp, language tag, domain classification, and token count estimates."}},{"@type":"Question",name:"Can Octoparse deliver data on a recurring schedule for RAG pipelines?",acceptedAnswer:{"@type":"Answer",text:"Yes. Recurring delivery on daily, weekly, or custom cadences is supported. Each batch is deduplicated against prior deliveries â only new or changed records are included â with consistent schema and freshness timestamps, so your retrieval pipeline always ingests clean, current data without manual intervention."}},{"@type":"Question",name:"What types of web data can be scoped for AI use cases?",acceptedA
1nswer:{"@type":"Answer",text:"Octoparse can scope datasets from product listings and reviews, news and editorial content, financial commentary, job postings, technical documentation, forum and community discussions, company profiles, social media platforms, and more. Coverage is confirmed per project after scoping â if a specific source or field is a hard requirement, that is verified before work begins."}},{"@type":"Question",name:"How does Octoparse handle deduplication and data quality?",acceptedAnswer:{"@type":"Answer",text:"Every delivery is processed for exact and near-duplicate removal, field normalization across sources, language detection, and structural QA. In typical domain corpus deliveries, deduplication reduces raw collection volume by 85â95%. For recurring deliveries, deduplication runs against previous batches so only genuinely new content reaches your pipeline."}},{"@type":"Question",name:"What is the licensing situation on web-collected data?",acceptedAnswer:{"@type":"Answer",text:"Octoparse collects publicly accessible web data â content openly available without authentication or paywall. Customers are responsible for confirming the legal basis for their intended use of collected data within their specific jurisdiction and application. Octoparse does not collect from paywalled, login-protected, or otherwise access-restricted sources."}},{"@type":"Question",name:"How fast can I get a sample dataset, and what does it cost?",acceptedAnswer:{"@type":"Answer",text:"A free scoped sample is typically delivered within 1â2 business days after you share your target domain, sources, and required fields â no commitment required to review the sample. Projects start from $699/project for one-time datasets and $599/month for recurring delivery. Large-scale or multi-domain requirements are scoped separately."}}]},_={"@context":"https://schema.org","@type":"BreadcrumbList",itemListElement:[{"@type":"ListItem",position:1,name:"Home",item:"https://www.octoparse.com/"},{"@type":"ListItem",position:2,name:"Data Service",item:"https://www.octoparse.com/data-service"},{"@type":"ListItem",position:3,name:"Web Data for AI",item:"https://www.octoparse.com/data-service/web-data-for-ai"}]},y=["Fine-tuning \xb7 RAG \xb7 Agent data feeds","JSONL \xb7 Parquet \xb7 S3 \xb7 BigQuery","Deduplicated \xb7 normalized \xb7 provenance-tagged","Free sample \xb7 No infra required"],x=[{value:"~94%",label:"dedup vs. raw crawl"},{value:"1â2d",label:"to first sample"},{value:"100%",label:"provenance-tagged"}],v=[{value:"1-2",suffix:"days",description:"Free sample dataset turnaround after scope confirmation - test quality before committing"},{value:"200+",description:"Source types and domains covered - from product reviews to financial content to technical forums"},{value:"99.9%",description:"SLA-backed delivery reliability for recurring dataset feeds and agent monitoring pipelines"},{prefix:"From",value:"$699",description:"Per dataset - $599/month recurring delivery - enterprise custom scope available"},{value:"4.8",suffix:"/5",description:"G2 rating - verified B2B data service reviews"},{value:"4.7",suffix:"/5",description:"Capterra rating - trusted by global data teams"}],j=[{icon:"training",title:"Fine-tuning & Training Data",description:"Domain-specific text corpora for LLM fine-tuning, continual pre-training, and RLHF datasets â scoped to your exact sources. Product reviews, forum discussions, news articles, technical documentation, industry content. Delivered deduplicated, normalized, and provenance-tagged with token count estimates included.",note:"Best for: domain adaptation, instruction tuning, evaluation dataset creation, RLHF preference data collection.",formats:["JSONL","Parquet","CSV","S3 / GCS"]},{icon:"rag",title:"RAG & Knowledge Base Feeds",description:"Recurring source-level content from your approved public sources â news, documentation, product pages, regulatory content, or competitor sites â delivered with timestamps, source URLs, and a consistent schema. Your team controls chunking, embeddings, indexing, retrieval, and evaluation.",n
1ote:"Best for: LLM applications needing fresh knowledge, enterprise knowledge bases, citation-grounded AI products.",formats:["JSONL","JSON","BigQuery","Snowflake"]},{icon:"agent",title:"AI Agent Data Feeds",description:"Structured, event-like signals for AI agents that need to monitor the real world â price changes, new product listings, competitor moves, job posting patterns, news events, regulatory filings. Delivered on a recurring schedule your agent can ingest directly, without building or maintaining a single collection workflow.",note:"Best for: autonomous agents, competitive intelligence systems, market monitoring AI, real-time pricing engines.",formats:["JSON","REST API","Webhook","DB push"]}],N=[{title:"Common Crawl",points:["The whole web â not your domain","Raw HTML requiring a major pipeline to parse and clean","Noise ratio makes training data quality unpredictable","Monthly release cadence â not aligned to your schedule"]},{title:"DIY Scraping",points:["Weeks to months of engineering before first usable dataset","Ongoing maintenance as source sites update or block","Your team owns dedup, normalization, and QA pipeline","Scales with headcount, not with data requirements"]},{title:"Bright Data / Apify",points:["Tools â you still write the collection and parsing logic","You build and maintain the normalization pipeline","Infra cost plus engineering time plus ongoing monitoring","No managed service â you operate it"]},{title:"Octoparse managed",points:["Domain-specific, scoped to your exact sources and fields","Structured, deduplicated, provenance-tagged output","JSONL / Parquet / warehouse-ready from day one","Recurring delivery â no infra for your team to maintain"],highlight:!0}],k=["LLM fine-tuning data","RAG knowledge base feeds","AI agent data pipelines","JSONL \xb7 Parquet \xb7 S3 \xb7 BigQuery","Deduplicated \xb7 provenance-tagged","Free sample \xb7 no commitment"],S=[{icon:"dedupe",title:"Deduplication",description:"Exact and near-duplicate detection applied across every batch. For recurring deliveries, deduplication runs against prior batches â only new or meaningfully changed records are included in each update, keeping your training corpus clean as it grows."},{icon:"provenance",title:"Provenance metadata",description:"Every record includes source URL, domain, publish timestamp, and language tag. Essential for RAG citation grounding, training data filtering by source credibility, and evaluation dataset construction with controlled source distribution."},{icon:"language",title:"Language detection",description:"Detected language and script are tagged at the record level â enabling locale filtering, multilingual training set composition, and per-market data separation without post-processing on your side."},{icon:"schema",title:"Schema consistency",description:"Field names, types, and formats are standardized across all sources in every delivery. One schema, regardless of how many source sites are in scope â so your ingestion pipeline handles one format, not one per source."},{icon:"token",title:"Token count estimates",description:"Approximate token counts are included per record â useful for training budget planning, context window management in RAG, and dataset composition decisions before fine-tuning runs begin. Basis: GPT-4 tokenizer."},{icon:"qa",title:"Pre-delivery QA",description:"Every delivery is reviewed for completeness, structural integrity, and field coverage before shipment. If there are gaps in scope coverage, we flag them rather than delivering partial data silently â and re-scope before proceeding."}],A=["id","source","url","domain","publish_time","lang","title","text","token_count","category","author","region"],L=[{field:"id",description:"Unique record identifier, stable across batches",example:"r_20260408_001"},{field:"source",description:"Source domain or publication name",example:"techcrunch.com"},{field:"url",description:"Canonical URL of the source content â provenance for RAG and evaluation",example:"https://techcrunch.com/2026/04/..."},{field:"domain",description:"Topic or industry classification label",example:"technology / ai / startups"},{field:"publish_time",description:"ISO 8601 UTC timestamp of original publication",example:"2026-04-08T09:14:00Z"},{field:"lang",description:"Detected language code (BCP 47)",example:"en / zh-CN / ja / ko"},{field:"text",description:"Extracted and cleaned body text, normalized encoding",example:"The funding round brings total..."},{field:"token_count",description:"Approximate token count (GPT-4 tokenizer basis)",example:"487"},{field:"category",description:"Source-assigned or inferred content category",example:"Funding / Product launch / Analysis"}],C=[{code:"JSONL",label:"LLM training standard"},{code:"Parquet",label:"large-scale processing"},{code:"CSV / JSON",label:""},{code:"S3 / GCS",label:"direct bucket push"},{code:"Snowflake \xb7 BigQuery \xb7 Redshift",label:""},{code:"REST API",label:"for agent ingestion"}],T=[{title:"RAG knowledge base for a legal AI product",description:"A legal tech startup building a contract analysis AI needed weekly updates of case law summaries, regulatory filings, and legal commentary â normalized, timestamped, and chunked for ingestion into a Pinecone vector store. Recurring delivery keeps their retrieval layer current without manual curation cycles."},{title:"E-commerce AI agent with live pricing signals",description:"An AI pricing agent needed
1structured, recurring price and inventory data across 50+ competitor product pages â delivered in JSON on a daily cadence, directly to the agent's ingestion API. No scraper to maintain; the agent focuses on decision logic, not data collection."},{title:"Multilingual training data for a global NLP model",description:"An NLP team needed domain-specific text in English, Chinese, Japanese, and Korean â language-tagged, schema-consistent, delivered in Parquet for large-scale preprocessing. Language detection and per-locale filtering enabled controlled dataset composition without additional post-processing."}],I=[{eyebrow:"Hard-source ecommerce dataset",title:"Temu pricing data pipeline case study for ecommerce intelligence",description:"Review how Octoparse operated a managed Temu pricing and inventory pipeline for a de-identified ecommerce intelligence client, delivering 8M+ monthly records in Phase 1, scaling toward 16M+ records, and shipping JSONL outputs to Snowflake with QA controls.",href:"/data-service/competitor-price-monitoring/temu-pricing-data-pipeline-case-study",cta:"Read the Temu pricing data pipeline case study",tags:["Temu pricing data","JSONL to Snowflake","workflow dataset"]},{eyebrow:"Public workflow dataset",title:"Multi-platform product matching case study for furniture and appliance retail",description:"See how Octoparse crawls public data from Wayfair, The Home Depot, Lowe's, Walmart, Target, and other retail sources, normalizes every platform into one product schema, and uses multi-signal plus visual matching to identify true product matches.",href:"/data-service/ai-visual-product-matching/retail-product-matching",cta:"Read the multi-platform product matching case study",tags:["retail product matching","normalized product schema","visual matching dataset"]},{eyebrow:"AI visual matching proof",title:"AI visual product matching workflow for noisy marketplace candidates",description:"Review a managed visual matching workflow that turns noisy candidate retrieval, pre-vision filtering, wrong-part rejection, visual scoring, and structured output buckets into an AI-ready workflow preview.",href:"/data-service/ai-visual-product-matching/automotive-parts-case-study",cta:"Read the AI visual product matching workflow",tags:["AI visual matching","pre-vision filtering","structured output"]}],O=[{value:"item-0",question:"What are custom web datasets for AI training?",answer:"Custom web datasets for AI training are scoped, structured datasets collected from approved public web sources for model training, LLM fine-tuning, RAG knowledge bases, AI agent monitoring, evaluation sets, or domain-specific model adaptation. Octoparse defines the sources, fields, schema, freshness cadence, provenance metadata, and delivery format with your team before collection begins."},{value:"item-1",question:"How is this different from Common Crawl?",answer:"Common Crawl is a broad monthly crawl of the entire web â useful as a pretraining base, but not domain-specific, and requiring a significant engineering effort to extract and clean the fraction relevant to your model. Octoparse delivers purpose-built datasets scoped exactly to your target domain, sources, and fields: structured, deduplicated, normalized, and delivered in JSONL or Parquet â not raw HTML that requires another pipeline project to make usable."},{value:"item-2",question:"What output formats are supported?",answer:"Octoparse delivers data as JSONL (the standard format for LLM fine-tuning workflows), Parquet (for large-scale data processing), CSV, JSON, or direct push to cloud storage â S3 or GCS â and data warehouses including Snowflake, BigQuery, and Redshift. Every record includes provenance metadata: source URL, timestamp, language tag, domain classification, and token count estimate."},{value:"item-3",question:"Can you deliver on a recurring schedule for our RAG pipeline?",answer:"Yes. Recurring delivery on daily, weekly, or custom cadences is fully supported. Each batch is deduplicated against prior deliveries â only new or changed records are included â and delivered with consistent schema and freshness timestamps. Your retrieval pipeline always ingests clean, current data without manual processing steps between deliveries."},{value:"item-4",question:"What domains and source types can be scoped?",answer:"Octoparse can scope datasets from product listings and reviews, news and editorial content, financial commentary and market data, job postings, technical documentation, forum and community discussions, company profiles, social media platforms, and more. Coverage is confirmed per project after scoping â if a specific source or field is a hard requirement, that's the first thing we verify before work begins."},{value:"item-5",question:"How does deduplication work, and how clean is the output?",answer:"Every delivery is processed for exact and near-duplicate removal, field normalization across sources, and structural QA. In typical domain corpus deliveries, deduplication reduces raw collection volume by 85â95%. For recurring deliveries, deduplication runs against previous batches so only genuinely new content reaches your ingestion pipeline."},{value:"item-6",question:"What's the licensing position on web-collected data?",answer:"Octoparse collects publicly accessible web data â content openly available without authentication or paywall. Customers are responsible for confirming the legal basis for their intended use within their specific jurisdiction and application type. Octoparse does not collect from paywalled, login-protected, or access-restricted sources."},{value:"item-7",question:"How fast can I get a sample, and what does it cost?",answer:"A free scoped sample is typically delivered within 1â2 business days after you share your target domain, sources, and required fields â no commitment needed before reviewing sample quality. Projects start from $699/project for one-time datasets and $599/month for recurring delivery. Large-scale or multi-domain requirements are scoped separately with a custom quote."},{value:"item-8",question:"How is this different from using Common Crawl, Bright Data, or a scraping API like Apify?",answer:"Common Crawl is a generic web archive â broad coverage, but no domain filtering, inconsistent schema, high noise, and n
1o freshness guarantees. Bright Data and Apify provide scraping infrastructure, but your team still designs the schema, builds the extraction logic, handles deduplication, and manages QA. Octoparse Managed delivers a purpose-built corpus: you specify the domains, content types, and schema fields you need; we handle extraction, deduplication (~94% reduction vs. raw crawl), provenance tagging, and format conversion to JSONL, Parquet, or direct push to S3, BigQuery, or Snowflake. The result is data that goes directly into your training or RAG pipeline â not into another cleaning backlog."},{value:"item-9",question:"What if my target domains block automated collection?",answer:"Anti-bot evasion is a core Octoparse engineering competency â we handle proxy rotation, JavaScript rendering, browser fingerprinting resistance, and rate-limit management as part of every managed project. Technical feasibility is assessed during the free scoping phase before any commitment. If a specific domain or content type isn't collectible under current conditions, we identify this upfront so you can adjust your corpus scope at no cost."},{value:"item-10",question:"Can I validate dataset quality before committing to a full production run?",answer:"Yes. Every project starts with a free scoped sample â delivered within 1â2 business days â covering your target domains with the full schema you requested. Sample records include all metadata fields (source URL, publish timestamp, language tag, domain classification, token count estimate) so your ML team can validate schema consistency, deduplication quality, and format compatibility with your training pipeline before approving a full corpus run."}],P=[{icon:"price",title:"Competitor Price Monitoring",description:"Need structured product and pricing data for your pricing model or retail AI? Our feeds are pre-matched, normalized, and schema-consistent.",href:"/data-service/competitor-price-monitoring"},{icon:"social",title:"Social Media Monitoring",description:"Training a sentiment or NLP model on social text? We deliver labelled, deduplicated content in JSONL from 60+ global platforms.",href:"/data-service/social-media-monitoring"},{icon:"lead",title:"B2B Lead Generation Data",description:"Building a company intelligence pipeline or ICP classifier? Our lead-gen data includes structured firmographic fields your model expects.",href:"/data-service/b2b-lead-generation"}],q=()=>{let e=(0,c.o)(),a=a=>{a.preventDefault(),e("WebDataForAI_RequestSample")},t=a=>{a.preventDefault(),e("WebDataForAI_TalkToExpert")};return(0,i.jsxs)(i.Fragment,{children:[(0,i.jsxs)(s.Z,{title:"Custom Web Datasets for AI Training, RAG & LLM Fine-Tuning | Octoparse",description:"Custom web datasets for AI training, RAG, LLM fine-tuning, and AI agent pipelines. Domain-specific, deduplicated, provenance-tagged, and delivered as JSONL, Parquet, CSV, Snowflake, BigQuery, S3, or API.",ogTitle:"Custom Web Datasets for AI Training | Octoparse",ogDescription:"Purpose-built structured web data for AI training, RAG pipelines, LLM fine-tuning, and agent monitoring workflows. Delivered as JSONL, Parquet, or direct warehouse sync.",twitterCard:"summary_large_image",twitterTitle:"Custom Web Datasets for AI Training | Octoparse",twitterDescription:"Domain-specific structured web data for LLM fine-tuning, RAG knowledge bases, and AI agent feeds. JSONL, Parquet, S3, BigQuery, Snowflake, and API.",children:[(0,i.jsx)("script",{type:"application/ld+json",dangerouslySetInnerHTML:{__html:JSON.stringify(g)}}),(0,i.jsx)("script",{type:"application/ld+json",dangerouslySetInnerHTML:{__html:JSON.stringify(w)}}),(0,i.jsx)("script",{type:"application/ld+json",dangerouslySetInnerHTML:{__html:JSON.stringify(b)}}),(0,i.jsx)("script",{type:"application/ld+json",dangerouslySetInnerHTML:{__html:JSON.stringify(_)}})]}),(0,i.jsxs)("div",{className:"".concat(h().page," ").concat(h().opTunedAiPage),children:[(0,i.jsx)("header",{className:"".concat(h().hero," ").concat(h().heroBlue),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsx)(n.Z,{}),(0,i.jsxs)("div",{className:h().heroGrid,children:[(0,i.jsxs)("div",{className:h().heroCopy,children:[(0,i.jsx)("span",{className:h().eyebrow,children:"Managed Data Service - Custom Web Datasets for AI Training"}),(0,i.jsxs)("h1",{className:"".concat(h().title," ").concat(h().titleOffset),children:[(0,i.jsx)("span",{className:h().titlePrimaryLine,children:"Custom Web Datasets"}),(0,i.jsx)("span",{className:h().titlePrimaryLine,children:"for AI Training,"}),(0,i.jsx)("span",{className:h().titlePurpleLine,children:"RAG, and LLM Fine-Tuning."})]}),(0,i.jsx)("p",{className:h().subtitle,children:"Structured, deduplicated, provenance-tagged public web datasets for LLM fine-tuning, RAG knowledge bases, AI agents, and model evaluation. Delivered as JSONL, Parquet, CSV, Snowflake, BigQuery, AWS S3, or API with
1no scraper to build and no pipeline to maintain."}),(0,i.jsx)("div",{className:h().proofRow,children:y.map(e=>(0,i.jsx)("span",{className:h().proofPill,children:e},e))}),(0,i.jsxs)("div",{className:h().actions,children:[(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnPrimary),onClick:a,children:"Get a Free Sample Dataset"}),(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnSecondary),onClick:t,children:"Talk to a Data Expert"})]}),(0,i.jsx)("p",{className:h().heroNote,children:"Common Crawl has everything. Your model needs the right 1%. Tell us your domain, target sources, fields, freshness cadence, and required format. Octoparse scopes the collection workflow, handles deduplication, schema normalization, QA, and delivery, then ships a production-ready source dataset for your team to validate, chunk, embed, and index in its own AI pipeline."})]}),(0,i.jsxs)("aside",{className:h().heroPanel,"aria-label":"Sample JSONL output",children:[(0,i.jsxs)("div",{className:h().panelTop,children:[(0,i.jsxs)("div",{children:[(0,i.jsxs)("p",{className:h().panelTitle,children:["Sample delivery ","â"," fine-tuning corpus"]}),(0,i.jsx)("p",{className:h().panelSub,children:"Finance domain \xb7 JSONL \xb7 deduplicated \xb7 provenance-tagged"})]}),(0,i.jsxs)("span",{className:h().panelBadge,children:[(0,i.jsx)(o.sf,{name:"jsonl",className:h().badgeIcon}),"JSONL"]})]}),(0,i.jsxs)("div",{className:h().codeBlock,children:[(0,i.jsxs)("div",{className:h().codeHead,children:[(0,i.jsxs)("div",{className:h().codeDots,children:[(0,i.jsx)("span",{}),(0,i.jsx)("span",{}),(0,i.jsx)("span",{})]}),(0,i.jsx)("span",{className:h().codeHeadLabel,children:"output.jsonl"})]}),(0,i.jsxs)("div",{className:h().codeBody,children:[(0,i.jsxs)("span",{className:h().cl,children:["{",(0,i.jsx)("span",{className:h().ck,children:'"id"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"r_001"'}),", ",(0,i.jsx)("span",{className:h().ck,children:'"source"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"seekingalpha.com"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"domain"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"finance"'}),", ",(0,i.jsx)("span",{className:h().ck,children:'"lang"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"en"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"publish_time"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"2026-04-08T14:22:00Z"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"token_count"'}),":",(0,i.jsx)("span",{className:h().cn,children:"487"}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"url"'}),":",(0,i.jsx)("span",{className:h().cu,children:'"https://seekingalpha.com/..."'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"text"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"Q3 earnings analysis..."'}),"}"]}),(0,i.jsxs)("span",{className:h().cl,children:["{",(0,i.jsx)("span",{className:h().ck,children:'"id"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"r_002"'}),", ",(0,i.jsx)("span",{className:h().ck,children:'"source"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"reddit.com/r/investing"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"domain"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"finance"'}),", ",(0,i.jsx)("span",{className:h().ck,children:'"lang"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"en"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"publish_time"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"2026-04-08T16:11:00Z"'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"token_count"'}),":",(0,i.jsx)("span",{className:h().cn,children:"312"}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"url"'}),":",(0,i.jsx)("span",{className:h().cu,children:'"https://reddit.com/r/investing/..."'}),","]}),(0,i.jsxs)("span",{className:h().cl,children:[" ",(0,i.jsx)("span",{className:h().ck,children:'"text"'}),":",(0,i.jsx)("span",{className:h().cs,children:'"Sector rotation thesis..."'}),"}"]})]})]}),(0,i.jsx)("div",{className:h().statRow,children:x.map(e=>(0,i.jsxs)("div",{className:h().statMini,children:[(0,i.jsx)("strong",{children:e.value}),(0,i.jsx)("span",{children:e.label})]},e.label))})]})]})]})}),(0,i.jsx)(r.Z,{items:v,containerClassName:h().container,ariaLabel:"Web data for AI service metrics"}),(0,i.jsx)("section",{className:"".concat(h().section," ").concat(h().sectionWhite),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight),children:"Three workflows, one service"}),(0,i.jsx)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide860," ").concat(h().sectionTitleOffset),children:"What AI teams actually use this for."}),(0,i.jsxs)("p",{className:h().sectionLead,children:["Fine-tuning a model, building a RAG application, or powering an AI agent ","â"," the bottleneck is always the same: clean, domain-specific, structured data that's actually ready to ingest. Generic crawls aren't it. DIY pipelines take months. This is the managed alternative."]})]}),(0,i.jsx)("div",{className:h().useCaseGrid,children:j.map(e=>(0,i.jsxs)("article",{className:h().useCaseCard,children:[(0,i.jsx)("div",{className:h().useCaseIcon,children:(0,i.jsx)(o.sf,{name:e.icon})}),(0,i.jsx)("h3",{children:e.title}),(0,i.jsx)("p",{children:e.description}),(0,i.jsx)("p",{className:h().useCaseNote,children:e.note}),(0,i.jsx)("div",{className:h().useCaseFormats,children:e.formats.map(e=>(0,i.jsx)("span",{className:h().useCaseFormat,children:e},e))})]},e.title))}),(0,i.jsxs)("aside",{className:h().answerGuide,"aria-labelledby":"rag-guide-title",children:[(0,i.jsxs)("div",{children:[(0,i.jsx)("span",{className:h().answerGuideEyebrow,children:"Practical guide + public sample"}),(0,i.jsx)("h3",{id:"rag-guide-title",children:"How to build a RAG-ready dataset from web data"}),(0,i.jsx)("p",{children:"Use the seven-step source-data workflow, inspect the field contract, and download 24 real records collected from public Octoparse documentation. The guide also marks exactly where Octop
1arse delivery ends and your RAG engineering begins."})]}),(0,i.jsxs)("div",{className:h().answerGuideAction,children:[(0,i.jsxs)(m(),{href:"/data-service/web-data-for-ai/how-to-build-a-rag-ready-dataset",className:"".concat(h().btn," ").concat(h().btnPrimary),children:["Open the RAG data guide",(0,i.jsx)(u.Z,{"aria-hidden":"true",size:17,strokeWidth:2})]}),(0,i.jsx)("span",{children:"JSONL, CSV, schema, QA, and methodology"})]})]}),(0,i.jsx)("div",{className:h().sectionHeader,children:(0,i.jsxs)("div",{className:"".concat(h().actions," ").concat(h().actionsOffset28),children:[(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnPrimary),onClick:a,children:"Request a Free Sample Dataset"}),(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnLight),onClick:t,children:"Discuss Your Use Case"})]})})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionWash," ").concat(h().sectionTop72),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight),children:"Why not the alternatives?"}),(0,i.jsx)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide820," ").concat(h().sectionTitleOffset),children:"Every AI team hits the same wall. Here's what the other paths actually cost."}),(0,i.jsx)("p",{className:h().sectionLead,children:"There are four common approaches to getting domain-specific web data. Three of them have hidden costs that most teams only discover after months of lost engineering time."})]}),(0,i.jsxs)("div",{className:h().diffBlock,children:[(0,i.jsxs)("h4",{children:["Octoparse vs. the alternatives ","â"," how they actually compare"]}),(0,i.jsx)("div",{className:h().diffGrid,children:N.map(e=>(0,i.jsxs)("div",{className:"".concat(h().diffCol," ").concat(e.highlight?h().diffColOcto:"").trim(),children:[(0,i.jsx)("h5",{children:e.title}),(0,i.jsx)("ul",{children:e.points.map(e=>(0,i.jsx)("li",{children:e},e))})]},e.title))})]})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionWhite," ").concat(h().sectionTop72),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().geoBlock,"aria-label":"About this service",children:[(0,i.jsxs)("div",{className:h().geoBlockKicker,children:[(0,i.jsx)(o.sf,{name:"sample",className:h().kickerIcon}),"About this service"]}),(0,i.jsxs)("p",{children:[(0,i.jsx)("strong",{children:"Octoparse Custom Web Datasets for AI Training"})," is a managed web data service that provides"," ",(0,i.jsx)("strong",{children:"domain-specific, structured datasets"})," for LLM fine-tuning, RAG knowledge bases, AI agents, and model evaluation. Teams define their target domains, sources, fields, freshness cadence, and required delivery format; Octoparse handles collection, deduplication, schema normalization, QA, provenance tagging, and structured delivery in ",(0,i.jsx)("strong",{children:"JSONL, Parquet, CSV"}),", or direct push to"," ",(0,i.jsx)("strong",{children:"S3, GCS, Snowflake, or BigQuery"}),". Unlike generic crawls such as Common Crawl, Octoparse delivers purpose-built corpora with full"," ",(0,i.jsx)("strong",{children:"provenance metadata"})," ","â"," source URL, publish timestamp, language tag, domain classification, and token count estimates"," ","â"," on a one-time or recurring schedule. Supported data types include product listings and reviews, news and editorial content, financial commentary, job postings, forum and community discussions, technical documentation, company profiles, and social media content. Projects start from ",(0,i.jsx)("strong",{children:"$699/project"})," for one-time datasets and ",(0,i.jsx)("strong",{children:"$599/month"})," for recurring feeds. Free scoped samples are delivered within ",(0,i.jsxs)("strong",{children:["1","â","2 business days"]}),"."]}),(0,i.jsx)("div",{className:h().geoPills,children:k.map(e=>(0,i.jsx)("span",{className:h().geoPill,children:e},e))})]}),(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight),children:"Data quality for AI"}),(0,i.jsx)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide820," ").concat(h().sectionTitleOffset),children:"The difference between data that trains a better model and data that corrupts one."}),(0,i.jsx)("p",{className:h().sectionLead,children:"Quality issues in training data compound during fine-tuning. Duplicates inflate model confidence on repeated patterns. Schema inconsistencies break ingestion pipelines. Missing provenance makes evaluation impossible. Octoparse applies a structured QA pipeline before every delivery."})]}),(0,i.jsx)("div",{className:h().qualityGrid,children:S.map(e=>(0,i.jsxs)("div",{className:h().qualityItem,children:[(0,i.jsx)("div",{className:h().qIcon,children:(0,i.jsx)(o.sf,{name:e.icon})}),(0,i.jsx)("h4",{children:e.title}),(0,i.jsx)("p",{children:e.description})]},e.title))})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionWash),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight," ").concat(h().eyebrowOffset40),children:"Delivered schema"}),(0,i.jsxs)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide820," ").concat(h().sectionTitleOffset),children:["Standard delivery schema ","â"," every field an AI pipeline needs, included by default."]}),(0,i.jsx)("p",{className:h().sectionLead,children:"Fields are confirmed at project scoping based on what is publicly accessible from your target sources. The schema below represents a standard text-content corpus delivery;
1 structured data (prices, listings, profiles) follows a separately defined schema per project."})]}),(0,i.jsx)("div",{className:h().chipRow,"aria-label":"Standard field examples",children:A.map(e=>(0,i.jsx)("span",{className:h().chip,children:e},e))}),(0,i.jsx)("div",{className:h().schemaBox,role:"region","aria-label":"Example delivery schema",children:(0,i.jsxs)("table",{children:[(0,i.jsx)("thead",{children:(0,i.jsxs)("tr",{children:[(0,i.jsx)("th",{children:"Field"}),(0,i.jsx)("th",{children:"Description"}),(0,i.jsx)("th",{children:"Example value"})]})}),(0,i.jsx)("tbody",{children:L.map(e=>(0,i.jsxs)("tr",{children:[(0,i.jsx)("td",{children:(0,i.jsx)("span",{className:h().fieldName,children:e.field})}),(0,i.jsx)("td",{children:e.description}),(0,i.jsx)("td",{children:e.example})]},e.field))})]})}),(0,i.jsx)("div",{className:h().formatRow,"aria-label":"Supported output formats",children:C.map(e=>(0,i.jsxs)("span",{className:h().formatPill,children:[(0,i.jsx)("code",{children:e.code}),e.label?" â ".concat(e.label):""]},"".concat(e.code,"-").concat(e.label)))}),(0,i.jsxs)("div",{className:h().enterpriseCallout,children:[(0,i.jsxs)("p",{children:[(0,i.jsx)("strong",{children:"Building at model scale?"})," For teams requiring multi-domain corpora, custom deduplication strategies, PII filtering pipelines, or integration with existing ML infrastructure ","â"," we scope these separately with a dedicated data expert."]}),(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnSecondary," ").concat(h().buttonMin210),onClick:t,children:"Discuss at Scale"})]})]})}),(0,i.jsx)("section",{className:"".concat(h().section," ").concat(h().sectionWhite),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight),children:"Representative workflow \xb7 one engagement of many"}),(0,i.jsxs)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide860," ").concat(h().sectionTitleOffset),children:["How AI teams get from data sourcing problem to first fine-tuning run ","â"," without a data engineering sprint."]})]}),(0,i.jsxs)("div",{className:h().featureCase,children:[(0,i.jsxs)("article",{className:h().caseCard,children:[(0,i.jsx)("div",{className:h().caseKicker,children:"Featured Case Study"}),(0,i.jsx)("h3",{className:h().caseTitle,children:"How a fintech AI startup built a domain-adapted LLM for investment research without hiring a data engineer"}),(0,i.jsxs)("div",{className:h().caseMeta,children:[(0,i.jsxs)("span",{className:h().caseTag,children:[(0,i.jsx)(o.sf,{name:"agent",className:h().tagIcon}),"LLM fine-tuning"]}),(0,i.jsx)("span",{className:h().caseTag,children:"Finance domain corpus"}),(0,i.jsx)("span",{className:h().caseTag,children:"JSONL \xb7 GCS delivery"})]}),(0,i.jsxs)("p",{className:"".concat(h().muted," ").concat(h().caseIntroText),children:["A fintech AI startup building a domain-adapted LLM for investment research needed a high-quality financial text corpus ","â"," earnings analysis, investor forum discussions, market commentary, and company news ","â"," structured, deduplicated, and ready for fine-tuning. Building their own collection pipeline was estimated at 2","â","3 months of engineering time and would require ongoing maintenance as source sites changed. Their ML team needed to focus on model work, not data infrastructure."]}),(0,i.jsxs)("div",{className:h().casePoints,children:[(0,i.jsxs)("div",{className:h().casePoint,children:[(0,i.jsx)("strong",{children:"Scope"}),(0,i.jsxs)("p",{children:["Initial one-time corpus of 500,000+ records across financial news sites, investment forums, and earnings-related content ","â"," JSONL format with source URL, timestamp, domain tag, and token count per record. Follow-on recurring weekly delivery of new content to keep the training corpus current over time."]})]}),(0,i.jsxs)("div",{className:h().casePoint,children:[(0,i.jsx)("strong",{children:"What Octop
1arse delivered"}),(0,i.jsxs)("p",{children:["First scoped sample dataset within 2 business days. Full corpus delivered in JSONL, deduplicated to ~94% of raw collection volume, with complete provenance metadata and consistent schema across all sources ","â"," pushed directly to the team's GCS bucket, ready for the fine-tuning pipeline without transformation steps."]})]}),(0,i.jsxs)("div",{className:h().casePoint,children:[(0,i.jsx)("strong",{children:"Business outcome"}),(0,i.jsxs)("p",{children:["The ML team went from data sourcing problem to first fine-tuning run in under two weeks ","â"," without building or maintaining a single scraper. Recurring weekly delivery keeps the corpus fresh without engineering involvement after initial setup."]})]})]}),(0,i.jsx)("div",{className:"".concat(h().actions," ").concat(h().actionsOffset24),children:(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnPrimary),onClick:a,children:"Start a Similar Project"})})]}),(0,i.jsx)("div",{className:h().caseSide,children:T.map(e=>(0,i.jsxs)("article",{className:h().miniProof,children:[(0,i.jsx)("span",{className:h().miniProofTag,children:"Use case"}),(0,i.jsx)("h3",{children:e.title}),(0,i.jsx)("p",{children:e.description})]},e.title))})]})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionGray),children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("span",{className:"".concat(h().eyebrow," ").concat(h().eyebrowLight),children:"Public workflow datasets & engineering case studies"}),(0,i.jsx)("h2",{className:"".concat(h().sectionTitle," ").concat(h().sectionTitleWide860," ").concat(h().sectionTitleOffset),children:"See how Octoparse turns messy public web data into AI-ready structured outputs."}),(0,i.jsx)("p",{className:h().sectionLead,children:"These case studies show the same data engineering principles AI teams need: hard-source collection, multi-source crawling, schema normalization, signal evaluation, visual matching, warehouse delivery, review buckets, and public workflow datasets for technical evaluation."})]}),(0,i.jsx)("div",{className:h().caseStudyGrid,children:I.map(e=>(0,i.jsxs)(m(),{href:e.href,className:h().caseStudyCard,children:[(0,i.jsx)("span",{className:h().caseStudyEyebrow,children:e.eyebrow}),(0,i.jsx)("h3",{children:e.title}),(0,i.jsx)("p",{children:e.description}),(0,i.jsx)("div",{className:h().caseStudyTags,children:e.tags.map(a=>(0,i.jsx)("span",{children:a},"".concat(e.title,"-").concat(a)))}),(0,i.jsx)("span",{className:h().caseStudyLink,children:e.cta})]},e.title))})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionGray),children:(0,i.jsxs)("div",{className:"".concat(h().container," mx-auto max-w-[900px]"),children:[(0,i.jsx)("div",{className:"mx-auto mb-4 w-max rounded-full border border-[#cfe1ff] bg-[#f0f7ff] px-[14px] py-2 text-[12px] font-bold leading-none text-[#3470ff]",children:"FAQ"}),(0,i.jsx)(d.Z,{faqList:O,compact:!0,title:"Questions AI and ML teams ask before their first dataset."})]})}),(0,i.jsx)("section",{className:h().relatedSection,children:(0,i.jsxs)("div",{className:h().container,children:[(0,i.jsxs)("div",{className:h().sectionHeader,children:[(0,i.jsx)("div",{className:h().relatedEyebrow,children:"Related Services"}),(0,i.jsx)("h2",{className:h().relatedTitle,children:"Complete Your Data Strategy"}),(0,i.jsxs)("p",{className:h().relatedSub,children:["Domain-specific web data is the foundation ","â"," enrich your AI pipeline with complementary structured feeds."]})]}),(0,i.jsx)("div",{className:h().relatedGrid,children:P.map(e=>(0,i.jsxs)(m(),{className:h().relCard,href:e.href,children:[(0,i.jsx)("div",{className:h().relIcon,children:(0,i.jsx)(o.Jr,{name:e.icon})}),(0,i.jsx)("p",{className:h().relName,children:e.title}),(0,i.jsx)("p",{className:h().relDesc,children:e.description}),(0,i.jsx)("span",{className:h().relLink,children:"View service"})]},e.title))})]})}),(0,i.jsx)("section",{className:"".concat(h().sectionTight," ").concat(h().sectionWhite),children:(0,i.jsx)("div",{className:h().container,children:(0,i.jsxs)("div",{className:h().finalBand,children:[(0,i.jsxs)("div",{children:[(0,i.jsxs)("h2",{children:["Your model's quality ceiling",(0,i.jsx)("br",{}),"is your data quality ceiling."]}),(0,i.jsxs)("p",{children:["Octoparse delivers domain-specific, structured web datasets for AI teams who need the right data ","â"," not the whole internet ","â"," ","ready to ingest without a data engineering project in between."]}),(0,i.jsx)("p",{className:h().pricingLine,children:"From $699/project \xb7 $599/month recurring \xb7 Enterprise custom scope \xb7 Free sample in 1â2 days"})]}),(0,i.jsxs)("div",{className:h().finalActions,children:[(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnWhite),onClick:a,children:"Get a Free Sample Dataset"}),(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnSecondary),onClick:t,children:"Talk to a Data Expert"})]})]})})}),(0,i.jsxs)("div",{className:h().mobileCta,"aria-label":"Mobile quick actions",children:[(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnPrimary),onClick:a,children:"Free Sample"}),(0,i.jsx)("button",{type:"button",className:"".concat(h().btn," ").concat(h().btnSecondary),onClick:t,children:"Talk to Expert"})]})]})]})};q.getLayout=l.G;var D=!0;a.default=q},75273:function(e){e.exports={bandWrap:"metrics-band_bandWrap__Uke1z",bandSurface:"metrics-band_bandSurface__GVH2M",grid:"metrics-band_grid__XhIKk",gridFour:"metrics-band_gridFour__W8Bbe",gridThree:"metrics-band_gridThree__4Nz0s",card:"metrics-band_card__KmsDy",measure:"metrics-band_measure__YvKzk",prefix:"metrics-band_prefix__VcwVG",value:"metrics-band_value__eX9Km",suffix:"metrics-band_suffix__2whTp",description:"metrics-band_description___v4AJ"}},94826:function(e){e.exports={faq:"faq_faq__m4uX2",faq_wrapper:"faq_faq_wrapper___c_Ef",title:"faq_title__XsJ4_",accordion:"faq_accordion__KyqE2",item:"faq_item__ny8hT",item_question:"faq_item_question__OdF0l",compact:"faq_compact__oj5xa",answer:"faq_answer__fHl5a",link:"faq_link__xjT_9"}},83487:function(e){e.exports={page:"web-data-for-ai_page__rwHF9",container:"web-data-for-ai_container__XRAtb",section:"web-data-for-ai_section__6zkuY",sectionTight:"web-data-for-ai_sectionTight__mL2r8",sectionWhite:"web-data-for-ai_sectionWhite__1m6cC",sectionGray:"web-data-for-ai_sectionGray__M2t3T",sectionHeader:"web-data-for-ai_sectionHeader__42eXc",eyebrow:"web-data-for-ai_eyebrow__e3DJb",sectionTitle:"web-data-for-ai_sectionTitle__cGE0J",sectionLead:"web-data-for-ai_sectionLead__5gvfV",actions:"web-data-for-ai_actions__YZdKT",eyebrowLight:"web-data-for-ai_eyebrowLight__iEPFk",title:"web-data-for-ai_title__mrHZx",t
1itlePrimaryLine:"web-data-for-ai_titlePrimaryLine__PM4bV",titlePurpleLine:"web-data-for-ai_titlePurpleLine__D6YbN",titleOffset:"web-data-for-ai_titleOffset__GjbHy",subtitle:"web-data-for-ai_subtitle__ZpMqO",sectionTitleOffset:"web-data-for-ai_sectionTitleOffset__PjFTD",sectionTitleWide860:"web-data-for-ai_sectionTitleWide860__eFiWO",sectionTitleWide820:"web-data-for-ai_sectionTitleWide820__p3yzu",muted:"web-data-for-ai_muted__mzbte",hero:"web-data-for-ai_hero__DY6YE",heroBlue:"web-data-for-ai_heroBlue__TU8eO",heroGrid:"web-data-for-ai_heroGrid__hayZl",heroCopy:"web-data-for-ai_heroCopy__UZW98",proofRow:"web-data-for-ai_proofRow__42g_3",proofPill:"web-data-for-ai_proofPill__xPM7z",actionsOffset28:"web-data-for-ai_actionsOffset28__p26og",actionsOffset24:"web-data-for-ai_actionsOffset24__jWLoi",btn:"web-data-for-ai_btn__PT_ri",btnPrimary:"web-data-for-ai_btnPrimary__yfCw8",btnSecondary:"web-data-for-ai_btnSecondary__5MZGP",btnLight:"web-data-for-ai_btnLight__J1K1c",btnWhite:"web-data-for-ai_btnWhite__8iwoT",answerGuide:"web-data-for-ai_answerGuide__WtUic",answerGuideEyebrow:"web-data-for-ai_answerGuideEyebrow__optM5",answerGuideAction:"web-data-for-ai_answerGuideAction__R266x",heroNote:"web-data-for-ai_heroNote__nhaQV",heroPanel:"web-data-for-ai_heroPanel__jGTi_",panelTop:"web-data-for-ai_panelTop__dbds4",panelTitle:"web-data-for-ai_panelTitle__sqcYt",panelSub:"web-data-for-ai_panelSub__u9ob7",panelBadge:"web-data-for-ai_panelBadge___4GUD",codeBlock:"web-data-for-ai_codeBlock__ipD9J",codeHead:"web-data-for-ai_codeHead___QnYw",codeHeadLabel:"web-data-for-ai_codeHeadLabel__77JS0",codeDots:"web-data-for-ai_codeDots__YapH_",codeBody:"web-data-for-ai_codeBody__IAuY5",cl:"web-data-for-ai_cl__GJdI7",ck:"web-data-for-ai_ck__dY_s2",cs:"web-data-for-ai_cs__kzwZh",cn:"web-data-for-ai_cn__0UwS4",cu:"web-data-for-ai_cu__MwjvI",statRow:"web-data-for-ai_statRow__u1ps1",statMini:"web-data-for-ai_statMini__77SYX",useCaseGrid:"web-data-for-ai_useCaseGrid__I5xHs",useCaseCard:"web-data-for-ai_useCaseCard__GM8rT",useCaseIcon:"web-data-for-ai_useCaseIcon__mnC_X",useCaseNote:"web-data-for-ai_useCaseNote___MwZi",useCaseFormats:"web-data-for-ai_useCaseFormats__tl46g",useCaseFormat:"web-data-for-ai_useCaseFormat__hxHSo",diffBlock:"web-data-for-ai_diffBlock__WnM6k",diffGrid:"web-data-for-ai_diffGrid__9PZRw",diffCol:"web-data-for-ai_diffCol__WbmlB",diffColOcto:"web-data-for-ai_diffColOcto__KjkQb",geoBlock:"web-data-for-ai_geoBlock___ahr4",geoBlockKicker:"web-data-for-ai_geoBlockKicker__y4KZP",geoPills:"web-data-for-ai_geoPills__Jstzm",geoPill:"web-data-for-ai_geoPill__kdqL0",qualityGrid:"web-data-for-ai_qualityGrid__gos6b",qualityItem:"web-data-for-ai_qualityItem__fWrMF",qIcon:"web-data-for-ai_qIcon__BzhBu",chipRow:"web-data-for-ai_chipRow__DyDGU",chip:"web-data-for-ai_chip__k9PCo",schemaBox:"web-data-for-ai_schemaBox__q9q8E",fieldName:"web-data-for-ai_fieldName__lB29X",formatRow:"web-data-for-ai_formatRow__TpsDl",formatPill:"web-data-for-ai_formatPill__l1uKg",enterpriseCallout:"web-data-for-ai_enterpriseCallout__cMuZE",featureCase:"web-data-for-ai_featureCase__7HFjP",caseCard:"web-data-for-ai_caseCard__i0ZTO",caseKicker:"web-data-for-ai_caseKicker__HNWMZ",caseTitle:"web-data-for-ai_caseTitle__X_Wa5",caseMeta:"web-data-for-ai_caseMeta__Ld1X6",caseTag:"web-data-for-ai_caseTag__2DHlN",casePoints:"web-data-for-ai_casePoints__PapuN",casePoint:"web-data-for-ai_casePoint__A4jZh",caseSide:"web-data-for-ai_caseSide__SntYT",miniProof:"web-data-for-ai_miniProof__t3hBB",miniProofTag:"web-data-for-ai_miniProofTag___xFYE",caseStudyGrid:"web-data-for-ai_caseStudyGrid__do1dX",caseStudyCard:"web-data-for-ai_caseStudyCard__UL16Z",caseStudyEyebrow:"web-data-for-ai_caseStudyEyebrow__DRUa0",caseStudyTags:"web-data-for-ai_caseStudyTags__rf2Y5",caseStudyLink:"web-data-for-ai_caseStudyLink__vbGEs",sectionWash:"web-data-for-ai_sectionWash__RZ_Az",sectionTop72:"web-data-for-ai_sectionTop72__pocBm",eyebrowOffset40:"web-data-for-ai_eyebrowOffset40__yAFO4",buttonMin210:"web-data-for-ai_buttonMin210___sHu7",caseIntroText:"web-data-for-ai_caseIntroText__yF8Z1",faqWrap:"web-data-for-ai_faqWrap__INOjk",faqList:"web-data-for-ai_faqList__B_2FL",faqListOffset26:"web-data-for-ai_faqListOffset26__BNCU5",faqAnswer:"web-data-for-ai_faqAnswer__GnecO",leadCard:"web-data-for-ai_leadCard__h3Vt6",leadTitleOffset:"web-data-for-ai_leadTitleOffset__xiHZr",leadDesc:"web-data-for-ai_leadDesc__oIwcJ",formGrid:"web-data-for-ai_formGrid__nnJWz",formRow:"web-data-for-ai_formRow__Wmuc_",field:"web-data-for-ai_field__CLWSH",btnSubmit:"web-data-for-ai_btnSubmit__E4lgN",formAssurance:"web-data-for-ai_formAssurance__TSfcX",formDivider:"web-data-for-ai_formDivider__GJf7g",helperList:"web-data-for-ai_helperList__ZV8SD",relatedSection:"web-data-for-ai_relatedSection__BnulG",relatedEyebrow:"web-data-for-ai_relatedEyebrow__d_6su",relatedTitle:"web-data-for-ai_relatedTitle__m9vuN",relatedSub:"web-data-for-ai_relatedSub__atKPo",relatedGrid:"web-data-for-ai_relatedGrid__N1yTK",relCard:"web-data-for-ai_relCard__lY67_",relIcon:"web-data-for-ai_relIcon__PeAHk",relName:"web-data-for-ai_relName__HJVBe",relDesc:"web-data-for-ai_relDesc__DrS2j",relLink:"web-data-for-ai_relLink__T7Nn9",finalBand:"web-data-for-ai_finalBand__MDEcg",pricingLine:"web-data-for-ai_pricingLine__qlEdk",finalActions:"web-data-for-ai_finalActions__8Sdk1",mobileCta:"web-data-for-ai_mobileCta__gg7Gy",opTunedAiPage:"web-data-for-ai_opTunedAiPage__XFaHg",badgeIcon:"web-data-for-ai_badgeIcon__svPtP",buttonIcon:"web-data-for-ai_buttonIcon__JgH60",kickerIcon:"web-data-for-ai_kickerIcon__Zsqrb",tagIcon:"web-data-for-ai_tagIcon__ghp8C"}},17019:function(e,a,t){"use strict";t.d(a,{Z:function(){return m}});var i=t(50959);/** 2 * @license lucide-react v1.39.0 - ISC 3 * 4 * This source code is licensed under the ISC license. 5 * See the LICENSE file in the root directory of this source tree. 6 */let s=(...e)=>e.filter((e,a,t)=>!!e&&""!==e.trim()&&t.indexOf(e)===a).join(" ").trim(),n=e=>e.replace(/([a-z0-9])([A-Z])/g,"$1-$2").toLowerCase(),r=e=>e.replace(/^([A-Z])|[\s-_]+(\w)/g,(e,a,t)=>t?t.toUpperCase():a.toLowerCase()),o=e=>{let a=r(e);return a.charAt(0).toUpperCase()+a.slice(1)};/** 7 * @license lucide-react v1.39.0 - ISC 8 * 9 * This source code is licensed under the ISC license. 10 * See the LICENSE file in the root directory of this source tree. 11 */var c={xmlns:"http://www.w3.org/2000/svg",width:24,height:24,viewBox:"0 0 24 24",fill:"none",stroke:"currentColor",strokeWidth:2,strokeLinecap:"round",strokeLinejoin:"round"};/** 12 * @license lucide-react v1.39.0 - ISC 13 * 14 * This source code is licensed under the ISC license. 15 * See the LICENSE file in the root directory of this source tree. 16 */let d=e=>{for(let a in e)if(a.startsWith("aria-")||"role"===a||"title"===a)return!0;return!1},l=(0,i.createContext)({}),u=()=>(0,i.useContext)(l),p=(0,i.forwardRef)(({color:e,size:a,strokeWidth:t,absoluteStrokeWidth:n,className:r="",children:o,iconNode:l,...p},m)=>
16{let{size:f=24,strokeWidth:h=2,absoluteStrokeWidth:g=!1,color:w="currentColor",className:b=""}=u()??{},_=n??g?24*Number(t??h)/Number(a??f):t??h;return(0,i.createElement)("svg",{ref:m,...c,width:a??f??c.width,height:a??f??c.height,stroke:e??w,strokeWidth:_,className:s("lucide",b,r),...!o&&!d(p)&&{"aria-hidden":"true"},...p},[...l.map(([e,a])=>(0,i.createElement)(e,a)),...Array.isArray(o)?o:[o]])}),m=(e,a)=>{let t=(0,i.forwardRef)(({className:t,...r},c)=>(0,i.createElement)(p,{ref:c,iconNode:a,className:s(`lucide-${n(o(e))}`,`lucide-${e}`,t),...r}));return t.displayName=o(e),t}},57402:function(e,a,t){"use strict";t.d(a,{Z:function(){return s}});var i=t(17019);let s=(0,i.Z)("arrow-right",[["path",{d:"M5 12h14",key:"1ays0h"}],["path",{d:"m12 5 7 7-7 7",key:"xquz4c"}]])}},function(e){e.O(0,[6841,5989,8827,312,8857,752,6919,9501,4125,9774,2888,179],function(){return e(e.s=54580)}),_N_E=e.O()}]);
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.