Floriane Le Floch

Autonomous AI-agent platform for marketing used by 1200+ B2B enterprises - Designed and shipped to production a multi-agent LLM system that runs end-to-end marketing workflows: market intelligence, content generation, analytics, publishing and continuous learning. - Built a knowledge-driven agent architecture (RAG + retrieval + memory) that learns each customer's brand context and improves from measured performance; removing prompt engineering on the user's side. - Sole full-stack engineer: agent orchestration (LangChain/LangGraph, multi-model - OpenAI, Gemini, Claude), vector-based retrieval & memory, Python/FastAPI services, Next.js frontend, AWS + Docker infra, observability (LangSmith, Sentry, PostHog), plus evals and guardrails.

I partner with teams to design and ship AI / generative-AI, data and Web3 products, from scoping to production. • Partner with founders, product and business teams to scope ambiguous problems and deliver working systems. • Ship full-stack across Python, TypeScript/Node, Next.js, SQL/NoSQL and AWS, plus LLM/GenAI integrations.

Blockchain infrastructure for circular commerce: resale & traceability for consumer brands. • Built the full technical stack from scratch: an abstraction layer over key management, gas fees and smart contracts so non-technical brands could launch resale/traceability programs without any blockchain literacy. • Drove customer adoption: 30+ brands onboarded and a New York Fashion Week launch partnership within the first year. • Invited speaker: XRPL APEX 2024 (Amsterdam) and the inaugural Women CTO Conference 2024.

First AI virtual assistant for project management. Applied ML/NLP deployed in production at Fortune 500 customers on billion-dollar industrial programs. • Designed and deployed an enterprise-grade AI into production for Fortune 500 customers in energy, defense and construction, owning the full loop: scoping, integration with client systems (SharePoint, Outlook), model training and production rollout. • Built predictive & prescriptive ML and NLP pipelines over terabytes of unstructured project data, surfacing risk before escalation and producing evidence used in high-stakes legal settlements. • Led product and delivery for Fortune 500 accounts (~$100k ACV)

AI assistant to reduce emergency-room waiting times. • Built an early applied-AI product end-to-end as a student founder. • 2nd prize, La Social Cup (France's national social-entrepreneurship competition).

Research work in innovative domains: OpenStack, SDN, Cloud, IPV6, Multicasting

- Developed automatic processes for repetitive tasks - Sketched a new internal application - Test new technologies