AI Engineer

Lemrock builds the infrastructure of agentic commerce. Tomorrow, consumers won't just buy on e-commerce websites. They will discover, compare and purchase directly inside conversational AI interfaces like ChatGPT, Perplexity, and a growing long tail of specialized AI assistants. For brands, this is a major shift: if they are not present in these environments, they become invisible.

enable any brand to exist, sell and distribute in this new channel. We raised a €7M seed round (the first in Europe on this market), with Jean-Baptiste Rudelle, founder of Criteo, on our board, and the Bpifrance DeepTech label. Today, Lemrock is already:

• 100M+ conversations per month
• 150+ brand clients
• a $100B market structuring at high speed CONTEXT You'll join a tight-knit team with high product standards and direct access to production deployment: our models are tested in real-world conditions, with several million users each month, with very short cycles between prototyping, integration and deployment. In this context, you'll work on the next generation of agentic and recommendation systems: building the pipelines, agents, and models that sit at the core of our product.

Your goal is to build and scale the agentic infrastructure that powers Lemrock's commerce intelligence: turning raw signals into automated, self-improving systems at production scale. Concretely, you will:

• Analyze large-scale conversational interaction datasets (100M+ events/month) to uncover behavioral patterns, intent signals, and performance drivers.
• Design and deploy agentic pipelines end-to-end, from data ingestion and enrichment to model orchestration, monitoring, and continuous improvement, integrated into systems exposed to millions of requests daily.
• Build autonomous agents that keep our knowledge infrastructure accurate and current.
• Translate insights into iteration loops in production, by updating, fine-tuning, and improving our existing recommendation and ranking algorithms with tight constraints on latency, robustness, and business outcomes.
• Design and train new recommendation models from scratch when needed, with a focus on scalability, evaluation rigor, and deployability in real-world traffic.

E

• Strong academic background (MVA, ENS, X, Central…)
• 2+ years of experience in AI, Agentic Systems, ML / Deep Learning, statistics or NLP
• Experience prototyping and deploying AI models into production
• Experience with production systems (APIs, monitoring, optimization)
• Strong interest in LLMs, recommendation and conversational systems: building agentic pipelines, agent orchestration, fine-tuning
• Comfortable with AI coding tools (Claude Code, Cursor)
• Thrives in ambiguity and 0-to-1 environments
• Fluent in English; French is a plus TECH STACK ML/AI (REQUIRED)
• Agentic frameworks (e.g., Langchain) and/or native SDKs like OpenAI/Anthropic
• Observability and evaluation for LLM/agent workflows
• Vector search, scoring, prompting, LLM orchestration
• Hybrid recommender systems, causal inference, probabilistic models BONUS (APPRECIATED BUT NOT REQUIRED)
• TypeScript
• Engineering: Docker, GCP, PostgreSQL, Redis, Vector DBs, PostHog -

Join a company in early breakout mode, already generating revenue, with a clear technological edge and strong financial backing to fuel its growth

• Work directly with an experienced team: two repeat YC founders who've scaled product & tech before, a former strategy/innovation director (10 years in the sector), and ex-strategy consultants (McKinsey QuantumBlack, BCG)
• Be part of the next structural shift of the web: agentic interfaces
• Own mission-critical topics, grow fast, and shape the company's future trajectory PACKAGE Competitive salary disclosed during the interview process.

The job is published on Lemrock’s website. Open the original posting ↗ to submit your application. Screeningpass does not fill in a form for you.

Start a contact search from this page. Screeningpass suggests up to three people connected to the role. Choose whom to write to, then Screeningpass looks for their professional email and prepares a message. You review it and send it yourself with Gmail. An email address is not always available, and a reply is never guaranteed.

A search costs 1 credit. Your credit is returned if the search fails. Signing up gives you 3 credits, enough for 3 searches.

Screeningpass brings together internships, apprenticeships and jobs published by companies. For a job, Screeningpass finds someone to contact at the company, looks for their professional email and prepares a message using your CV. See how it works in detail.