PageSourceSearch

https://flex.ai/assets/CaseStudies-B_6yigro.js

js flex.ai collected 2026-09-25 04:11:53 UTC 5,995 bytes, 1 lines download raw bytes

1import{d as e,n as t}from"./react-vendor-Bx0rRGkx.js";import{a as n,d as r,n as i,t as a}from"./SEOHead-D2_Na5KX.js";import{t as o}from"./arrow-right-DFnLw3cV.js";import{n as s}from"./seo-CCvDaH4G.js";var c=e(),l=[{slug:`legml`,kicker:`Legal AI`,title:`LegML + FlexAI: Ready for French Business Law`,summary:`LegML fine-tuned a 32B-parameter legal LLM that outperformed a frontier model on domain benchmarks, cutting both training and serving costs with FlexAI.`,stats:[{value:`+10%`,label:`Factual precision`},{value:`14 days`,label:`Training time`},{value:`4×`,label:`Lower serving cost`}]},{slug:`dragonllm`,kicker:`Financial AI`,title:`DragonLLM × FlexAI: Sovereign AI for Finance`,summary:`DragonLLM deployed fine-tuned financial models on FlexAI's sovereign infrastructure in France, with managed autoscaling, scale-to-zero, and zero operational surprises.`,stats:[{value:`100%`,label:`Sovereign hosting`},{value:`0 → ∞`,label:`Autoscaling`},{value:`>99.9%`,label:`Uptime`}]},{slug:`pixelcut`,kicker:`Image AI`,title:`Pixelcut × FlexAI: Simple Fine-Tuning for Image AI`,summary:`Pixelcut fine-tuned their image generation model on FlexAI's platform, choosing it for usage-based pricing, operational simplicity, and fast time to production.`,stats:[{value:`Pay-per-use`,label:`Cost model`},{value:`Plug & play`,label:`Simplicity`},{value:`Image gen`,label:`Model type`}]}],u=()=>(0,c.jsxs)(`div`,{className:`min-h-screen bg-background noise-overlay`,children:[(0,c.jsx)(a,{title:`AI Infrastructure Case Studies and Success Stories | FlexAI`,description:`How AI companies build on FlexAI: LegML cut compute costs 75%, DragonLLM shipped in-country inference at >99.9% uptime, Pixelcut moved to pay-per-use fine-tuning.`,canonical:`/case-studies`,jsonLd:s({name:`FlexAI Customer Case Studies`,description:`Customer stories from teams running training, fine-tuning, and inference on FlexAI's managed AI cloud.`,path:`/case-studies`,items:l.map(e=>({name:e.title,url:`/case-studies/${e.slug}`}))})}),(0,c.jsx)(n,{}),(0,c.jsxs)(`div`,{className:`pointer-events-none fixed inset-0 overflow-hidden`,children:[(0,c.jsx)(`div`,{className:`absolute -top-56 left-1/2 h-136 w-136 -translate-x-1/2 rounded-full bg-primary/10 blur-3xl`}),(0,c.jsx)(`div`,{className:`absolute bottom-0 right-0 h-120 w-120 rounded-full bg-secondary/20 blur-3xl`}),(0,c.jsx)(`div`,{className:`absolute inset-0 grid-background opacity-40`})]}),(0,c.jsxs)(`main`,{id:`main`,className:`relative pt-24`,children:[(0,c.jsx)(`section`,{className:`mx-auto max-w-6xl px-6 py-16 text-center`,children:(0,c.jsxs)(r.div,{initial:{y:20},animate:{y:0},transition:{duration:.5},children:[(0,c.jsx)(`div`,{className:`text-xs font-semibold tracking-widest text-primary/80 uppercase mb-3`,children:`Resources`}),(0,c.jsxs)(`h1`,{className:`font-display text-4xl md:text-6xl font-bold tracking-tight mb-4`,children:[`Case `,(0,c.jsx)(`span`,{className:`text-gradient glow-text`,children:`Studies`})]}),(0,c.jsx)(`p`,{className:`mx-auto max-w-2xl text-base leading-7 text-muted-foreground sm:text-lg`,children:`Real-world outcomes from production teams building with FlexAI.`})]})}),(0,c.jsxs)(`section`,{className:`mx-auto max-w-4xl px-6 pb-24`,children:[(0,c.jsx)(`div`,{className:`space-y-6`,children:l.map((e,n)=>(0,c.jsx)(r.div,{initial:{opacity:0,y:20},whileInView:{opacity:1,y:0},viewport:{once:!0},transition:{delay:n*.1},children:(0,c.jsxs)(t,{to:`/case-studies/${e.slug}`,className:`block rounded-2xl border border-border bg-card p-8 hover:border-primary/30 transition-all group`,children:[(0,c.jsx)(`div`,{className:`text-xs font-semibold tracking-widest text-primary/80 uppercase mb-2`,children:e.kicker}),(0,c.jsx)(`h2`,{className:`font-display text-xl sm:text-2xl font-bold mb-3 group-hover:text-primary transition-colors`,children:e.title}),(0,c.jsx)(`p`,{className:`text-sm leading-6 text-muted-foreground mb-6`,children:e.summary}),(0,c.jsx)(`div`,{className:`flex flex-wrap gap-6 mb-4`,children:e.stats.map(e=>(0,c.jsxs)(`div`,{children:[(0,c.jsx)(`div`,{className:`font-display text-lg font-bold text-primary`,children:e.value}),(0,c.jsx)(`div`,{className:`text-xs text-muted-foreground`,children:e.label})]},e.label))}),(0,c.jsxs)(`span`,{className:`inline-flex items-center gap-1 text-sm font-semibold text-primary group-hover:gap-2 transition-all`,children:[`Read case study `,(0,c.jsx)(o,{className:`w-4 h-4`})]})]})},e.slug))}),(0,c.jsx)(`div`,{className:`mt-16 overflow-x-auto rounded-2xl border border-border bg-card/50`,children:(0,c.jsxs)(`table`,{className:`w-full text-left text-sm`,children:[(0,c.jsx)(`caption`,{className:`px-6 pt-6 text-left text-xs text-muted-foreground`,children:`Published outcomes by customer. Each row links to the full case study, where the scope behind that figure is stated.`}),(0,c.jsx)(`thead`,{children:(0,c.jsxs)(`tr`,{className:`border-b border-border text-muted-foreground`,children:[(0,c.jsx)(`th`,{scope:`col`,className:`px-6 py-3 text-xs font-semibold uppercase tracking-wider`,children:`Customer`}),(0,c.jsx)(`th`,{scope:`col`,className:`px-6 py-3 text-xs font-semibold uppercase tracking-wider`,children:`Industry`}),(0,c.jsx)(`th`,{scope:`col`,className:`px-6 py-3 text-xs font-semibold uppercase tracking-wider`,colSpan:3,children:`Published outcomes`})]})}),(0,c.jsx)(`tbody`,{children:l.map(e=>(0,c.jsxs)(`tr`,{className:`border-b border-border/50 last:border-0`,children:[(0,c.jsx)(`th`,{scope:`row`,className:`px-6 py-3 font-semibold`,children:(0,c.jsx)(t,{to:`/case-studies/${e.slug}`,className:`text-primary hover:underline`,children:e.title.split(/\s[+\u00d7]\s/)[0]})}),(0,c.jsx)(`td`,{className:`px-6 py-3 text-muted-foreground`,children:e.kicker}),e.stats.map(e=>(0,c.jsxs)(`td`,{className:`px-6 py-3 text-foreground`,children:[(0,c.jsx)(`span`,{className:`font-semibold`,children:e.value}),` `,(0,c.jsx)(`span`,{className:`text-muted-foreground`,children:e.label.toLowerCase()})]},e.label))]},e.slug))})]})})]})]}),(0,c.jsx)(i,{})]});
1export{u as default};

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.