1import{j as e}from"./index-BZ4gXfUu.js";import{P as t}from"./PillarPage-sPb9h3QK.js";import{Z as a}from"./zap-oFnYBBf8.js";import"./site-BcZyatrh.js";import"./AiParticleCluster-CwTlCJKn.js";const l=()=>e.jsx(t,{Icon:a,eyebrow:"AI/ML Services",title:"Generative AI",description:"From GenAI strategy to production. We build large language model and multimodal solutions tailored to real enterprise workflows, with the guardrails to ship them safely.",summary:"Strong retrieval patterns, evaluation pipelines, and guardrails keep generative outputs accurate, grounded, and auditable across copilots, content, and decision support.",summaryLabel:"What we deliver",backTo:"/ai-ml-services",backLabel:"Back to AI/ML Services",groups:[{heading:"GenAI Strategy & Use Cases",tasks:["Identify high-leverage GenAI opportunities","Build vs buy vs fine-tune decisions","Prioritize by value, risk, and time to impact","Define measurable success metrics"]},{heading:"RAG & Knowledge Systems",tasks:["Retrieval-augmented generation over enterprise data","Vector, hybrid, and graph-based retrieval","Document parsing, chunking, and metadata strategies","Permission-aware grounding and citations"]},{heading:"Copilots & Multimodal Apps",tasks:["Domain copilots for analysts, engineers, and operators","Text, image, audio, and document workflows","Tool use, function calling, and structured outputs","Seamless integration with existing enterprise UX"]},{heading:"Evaluation & Guardrails",tasks:["Automated eval suites for accuracy and grounding","Bias, toxicity, and PII detection","Prompt injection and jailbreak resistance","Cost, latency, and quality monitoring in production"]}],ctaHeadline:"GenAI without governance is a liability. With it, a multiplier.",ctaSubtext:"We deploy generative AI tied to risk controls, evaluation, and human-in-the-loop oversight: productivity that scales without exposing the enterprise."});export{l as component};
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.