About the Role:

We are looking for an AI Engineer with around two years of experience to build and ship generative AI features into production. You will design LLM-powered workflows, use AI coding tools to move fast without shipping fragile code, and make sure what we release is safe, explainable, and defensible. Strong programming fundamentals come first: AI tools amplify good engineers, they do not replace the judgment needed to review, debug, and own what goes to production.

What You'll Do:

Generative AI:

- Build and ship GenAI features - RAG pipelines, agents, tool-calling workflows, structured extraction, summarization.

- Design, version, and test prompts; treat prompt changes as code changes with review.

- Tune retrieval - embeddings, chunking, vector search, re-ranking - because retrieval quality decides output quality.

- Integrate LLM APIs into backend services with proper timeouts, retries, fallbacks, and cost controls.

- Build evaluation harnesses: quality metrics, curated test sets, regression checks before every release.

AI-Assisted Development:

- Use AI coding tools (Claude Code, Cursor, Copilot, or similar) daily to prototype, refactor, and ship faster.

- Write precise specs and context for those tools - good output starts with a well-framed problem.

- Review AI-generated code critically: hallucinated APIs, silent logic errors, missing edge cases, security gaps.

- Own everything you commit, regardless of how it was produced; share effective practices with the team.

AI Governance & Responsible AI:

- Build guardrails into features: input validation, output filtering, PII handling, prompt-injection defenses, human-in-the-loop checkpoints.

- Maintain traceability - logging, prompt and model versioning, audit trails for AI decisions.

- Document model choices, known limitations, failure modes, and intended use for each AI feature.

- Test for bias, unsafe output, and data leakage as a standard part of the release process.

- Partner with legal, security, and compliance to keep systems aligned with internal policy and regulation.

Engineering & Cloud:

- Write clean, tested, documented production code and participate actively in code review.

- Deploy and run containerized services on cloud infrastructure with monitoring, alerting, and secret management.

- Work with product and design to scope what is realistically achievable with today's models.

What We're Looking For:

Required:

- 2+ years of software engineering experience, including hands-on work on GenAI or LLM-based systems.

- Strong programming skills - deep Python fluency plus at least one other language (Go, Java, TypeScript, C++).

- Something real you shipped and maintained on LLM APIs (OpenAI, Anthropic, or open-weight models), not just tutorials.

- Working knowledge of RAG: embeddings, chunking, vector search, and retrieval evaluation.

- Fluent daily use of AI coding assistants, with the judgment to review and correct their output.

- Awareness of AI risk - prompt injection, hallucination, data leakage, bias - and how to mitigate each in a real system.

- Basic cloud competence on AWS, GCP, or Azure: deploy a containerized service, manage secrets, read logs and metrics.

- Git, Docker, and CI/CD fluency, plus clear written communication about design decisions and tradeoffs.

Nice to Have:

- Orchestration or agent frameworks: LangChain, LlamaIndex, DSPy, tool-calling agents.

- Vector databases: pgvector, Pinecone, Qdrant, Weaviate, or FAISS.

- Evaluation and observability tooling: LangSmith, Langfuse, Weights & Biases, MLflow.

- Fine-tuning experience (LoRA, QLoRA, SFT) with Hugging Face Transformers.

- Security background - threat modeling, secure coding, or red-teaming AI systems.

- Open-source contributions or public projects we can look at