Antoine Krychowski

I design and deploy AI tools (agents, RAG, LLM automations) for teams that want a concrete deliverable in production.

My missions revolve around three axes. First, automating repetitive tasks with an LLM: extracting and classifying documents (PDFs, emails, forms), automatically generating reports, processing customer requests. Second, building an AI assistant or agent connected to your internal data: RAG on your documentation, an agent that performs actions in your tools, a specialized copilot for your teams. Finally, auditing or optimizing an existing AI project: prompt quality, API costs, latency, model selection, observability.

My background: Master's degree in Financial Engineering at ESILV, interning in NLP/AI at Moneta Asset Management until July 2026, where I am designing an LLM pipeline for analyzing financial documents of listed companies.

Last summer, as a Data/AI intern at a Parisian startup (OctaEdge), I built an LLM extraction pipeline on AWS that transforms business PDFs into a queryable database. In parallel, I am deploying AI automations in a construction SME, where I learned to build tools that operational staff actually use.

Mission format: I target short projects (1 to 5 days), with a pre-defined deliverable and async communication, typically an agent prototype, an LLM pipeline audit, or a RAG pilot. I work evenings and weekends during my internship, which suits this format well. Extended availability from August 2026.

My added value compared to a purely engineering profile: I am comfortable discussing with your business teams, understanding your problem in its context, defining the right scope, and delivering a tool that your teams adopt.

To start simple: 30 minutes via video call, you describe your need, I tell you if it's feasible, within what timeframe and at what cost.