AI Engineer

Balt; Paris; Sur place; EUR 70 000 - 120 000; Plein temps

Résumé du poste

Balt is building an AI employee that lives inside Slack and Teams for consulting firms and recruitment agencies, already in production with clients. We seek a hands-on engineer to own the agent, its routines, tool-calling, and long-running workflows.

You will ship integrations with clients' tools (Slack, Teams, ATS/CRM, job boards), advance document generation fidelity, and optimize model routing across providers while contributing to CI/CD and monitoring.

Qualifications

• Proficient in Python with async design and API-heavy backends.
• Hands-on with LLMs, tool usage, and prompt/context engineering.
• Experience with cloud environments (Azure, AWS, Modal, Vercel).
• Commitment to clean code, testing and observability.

Responsabilités

• Build and improve Balt's agent: routines, skills, tool-calling and long-running workflows.
• Ship integrations with Slack, Teams, ATS/CRM, job boards and MCP servers.
• Improve document generation quality: consultant profiles and RFP templates.
• Optimize model routing, cost and latency across providers (Anthropic, OpenAI, Mistral).
• Contribute to evals, monitoring and CI/CD around agent reliability.
• Collaborate directly with the founder and users to ship fast and iterate on feedback.

Connaissances

Python

Async programming

LLMs

Tool use

Prompt engineering

Outils

Azure

AWS

Modal

Vercel

Description du poste

We're building Balt, an AI employee in Slack & Teams for consulting firms and recruitment agencies. Already live with clients.

https://www.hire-balt.com

Your mission:

• Build and improve our agent: routines, skills, tool-calling and long-running workflows
• Ship integrations with the tools our clients already use (Slack, Teams, ATS/CRM, job boards, MCP servers)
• Work on document generation quality (consultant profiles, RFP responses) — templates, formatting, fidelity
• Improve model routing, cost and latency across providers (Anthropic, OpenAI, Mistral)
• Contribute to evals, monitoring and CI/CD around agent reliability
• Work directly with the founder and users, ship fast, iterate on real feedback

Your profile:

• Strong Python, comfortable with async and API-heavy backends
• Hands-on with LLMs: tool use, RAG, MCP, prompt and context engineering
• Familiar with cloud environments (Azure, AWS, Modal, Vercel...)
• You care about clean code, tests and observability
• Startup experience is a big plus but not required
• Early-stage startup, clients in production, public launch this month
• Ownership from day one, direct impact on the roadmap