The independent studio that designs, ships, and operates production AI. Twenty-two live products across intelligence, content, infrastructure, knowledge, compliance, and consumer — built for teams that need software that works on day one and still works on day one thousand.
PhoLabs designs and operates intelligence platforms, content engines, MCP infrastructure, and knowledge tools under one opinionated stack. Same architecture, same conventions, same standards across every product we touch.
Twelve hours of architecture for every minute of interface. Every product has a live domain, real users, and operating metrics — no demos that won't deploy, no prototypes without a path to revenue.
For teams that need bespoke systems rather than off-the-shelf products, our consulting branch — ShowdownLab — engineers custom AI for digital marketing, e-commerce, luxury, and trading desks. Same studio, same standards, different menu.
See all 22 products →
60+ live world feeds ingested in real time. Threat, conflict, sentiment, and market-impact analysis — plotted on an interactive geo-map and streamed to a Claude Opus analyst chat. Built for desks that need decision-grade intel, not headlines.
Visit Argus ↗
YouTube, podcasts, PDFs, course material — distilled into layered, queryable knowledge. Knowledge graphs, chronological walkthroughs, automatic glossary extraction. Ten hours of source becomes 500 pages of structured, navigable notes.
Visit KondensMe ↗
440,000 contacts ingested in under a minute. Natural-language segmentation, K-means clustering, multichannel campaigns, A/B testing, autopilot. Built for growth teams that need to ship at scale without an MMP duct-taped to a CDP.
Visit Camoni ↗
Image to full campaign in six orchestrated nodes: audit, five-persona brainstorm, production factory, viral scoring, video output via Seedance 1.5 Pro. React Router v7, Clerk auth, graph integration — production-ready out of the box.
Visit Lumina ↗
From real-time intelligence to consumer anti-spam, every product PhoLabs ships draws from one of these capability stacks. Same conventions, same deploy pipeline, same team — pick the discipline that matches your problem.
Every PhoLabs product has a live domain, a deploy pipeline, and real users. We don't ship demos that won't deploy, prototypes that won't scale, or features that aren't measured.
Real-time geopolitical intelligence. Automated sports content engines. MCP gateway infrastructure. F&B ingredient catalogs. Plus the bespoke client work shipped through our consulting branch ShowdownLab — for digital marketing, e-commerce, luxury, and trading desks. All built on one opinionated stack, so the next product ships faster than the last.
See all 22 products →
Fully automated French football publication. Articles, images, SEO, newsletter — end to end, zero editorial intervention. Live since 2024.
Real-time facial recognition for connected eyewear. Stream, identify, and alert in under a second — pushed straight to operator dashboards.
WhatsApp-native AI assistant for non-technical users — strategic planning, content creation, knowledge access. Five-package monorepo on Fastify, WebSocket, React Native, FastAPI, Next.js.
Vector search for ingredients — by name, by season, or by feeling. Query "clear sky" or "early autumn" and get matches. Next.js 15.3 + custom embeddings.
One MCP endpoint, every internal tool exposed. Unified auth, rate-limiting, and routing — the connective tissue across all PhoLabs products in production.
A structured engagement that gets your product live in weeks, not quarters. No staging-forever, no demos-without-deploys, no scope creep.
You share the problem. We pressure-test the thesis, the audience, and the moat. If we don't believe in it, we tell you — and we recommend who would.
We design the system end-to-end: data model, deploy target, observability, and the path to first paying user. You get a written plan, a fixed quote, and a deadline.
We ship to a real domain with real auth, real data, and real money flow. Daily progress logs, weekly demos, async-first communication. No surprises.
We run it for you, hand off to your team with documentation, or spin it out as a co-owned product. Either way, the product keeps shipping after we leave.
Pick the format that matches your stage. We're transparent about price, scope, and outcomes from the first call.
A focused build to validate a thesis or ship a single product. You bring the spec; we bring the team and the deploy. Best for teams that need a working prototype on a real domain — fast.
An embedded engagement. We co-build with your team, own the architecture, and ship continuously. Best for teams that need ongoing AI capacity without hiring a full studio internally.
We become your CTO and engineering team for one vertical. We own the build, the deploy, and a meaningful slice of the upside. Best for founders who need a technical co-founder who has shipped 22 products before.
Build Sprints are fixed-price and fixed-scope, quoted after the architecture phase. Partnerships are monthly retainers with outcomes negotiated quarterly. Co-founder engagements are equity-led with a base retainer. We share quotes within three business days of intake — no salesfunnel, no markups.
Defaults: TypeScript, React, Next.js, FastAPI, Python, Postgres, Redis, Cloudflare, Vercel, Fly.io. AI providers: Anthropic Claude, OpenAI, Google. Vector DBs: Qdrant, Pinecone. We pick the stack per problem, not by religion — and we don't onboard frameworks we can't operate.
Yes. Partnership and co-founder engagements both include knowledge transfer and pair-programming with your engineers. We don't build black boxes — your team should be able to own the system after we leave.
You own everything we build for you. Build Sprints transfer full IP at handoff. Partnerships and co-founder engagements operate under custom agreements that match the deal structure — and we sign your NDA, not ours.
Three options, decided at engagement start: we run it for you (ops retainer), we hand it off to your team (with documentation and 30-day support), or we spin it out as a co-owned product. The choice is yours.
Yes — but only for co-founder engagements. We're selective. We need to believe the thesis, the market, and the founder. Send your deck and we'll respond within three business days, even if it's a no.
Paris, France. Async-first, deep-work culture. We work with teams across Europe and North America on overlapping hours. No status meetings, no Slack-as-deliverable — just code, demos, and clear writing.
Founders and product leaders who know what they want, can decide quickly, and care about the craft. We're a poor fit for teams that need an agency to make decisions for them. We're an excellent fit for teams that have a thesis and need a studio to execute it.
We take on a small number of partner engagements per year. If your thesis resembles one of the twenty-two above, send it over — we reply within one business day, even if it's a no.
