First seen by Alion on Nov 17, 2026.
About the role
We are looking for a Senior Python AI Engineer to join a senior team that builds GenAI systems for a US client. The systems run in production and support domain experts in their daily work. The team uses coding agents as its standard development workflow, evaluates changes on data before release and works with a high degree of independence.
Projects
Agentic decision support for document review. The system processes submissions that consist of scanned PDFs, forms, spreadsheets and archives. It extracts structured data, applies review rules implemented as code or as LLM agents, and prepares a decision draft that a domain expert approves or revises. The work covers document extraction with traceability to source pages, the split between deterministic logic and LLM components, handling of model nondeterminism, and an evaluation harness that compares system output with expert decisions.
Question answering service. A retrieval and answer API used by several applications, including a public-facing one. It includes hybrid search, reranking, citations, guardrails, and ingestion pipelines for heterogeneous sources.
Research platform. A full-stack application with LLM features in an editorial workflow: source library, deduplication, AI-assisted drafting, review, and approval.
Smaller projects. Custom tools on an internal LLM platform, data pipelines and market intelligence analytics.
Our expectations
1. Core engineering
- 5+ years of experience in software development, mostly in Python
- Backend services and REST APIs in FastAPI, Flask or Django
- Automated testing with pytest
- Type hints, static analysis and linting (mypy, ruff), CI
- Ability to work across the stack, including frontend changes
2. Agentic development
- Daily use of coding agents (Claude Code, Cursor, Codex or similar)
- Writing task specifications for agents and reviewing their output with tests and data
- Configuring and maintaining the agent harness: project instructions, skills, hooks, subagents and MCP servers
3. GenAI and LLM
- 2+ years building LLM applications used in production
- Tool calling and multi-step agent workflows
- Combining rule-based logic with LLM components
- Structured outputs (Pydantic or equivalent)
- RAG: embeddings, vector and hybrid search, reranking
4. Evaluation
- Designing evaluations on labelled data and tracking regressions
- LLM observability, including cost and latency monitoring
5. Cloud
- Experience with AWS, Azure or GCP, including managed AI services
6.Soft Skills
- Independent work and ownership of a project or area
- Direct work with client domain experts and translating requirements into technical solutions
- Communicating technical decisions to non-technical stakeholders
- Fluent written and spoken English
- Remote work with daily overlap with US time zones (CET afternoon), usually 9-5
Welcome Skills
- Document processing, OCR, and structured data extraction
- AWS Bedrock
- Pydantic AI or LangGraph
- Langfuse or similar observability tools
- PostgreSQL and pgvector
- Next.js, TypeScript and Playwright
- Experience in consulting or client projects
Tech stack
- Python
- REST API
- LLM
- RAG
- Claude Code
- Cursor
- Codex
- AWS
- Azure
- GCP

