Senior AI Engineer

What You’ll Do

1. AI feature design & integration into product flows
   • Own AI feature requirements jointly with Product
   • Translate procurement logic into structured prompts, schemas, etc
   • Work closely with other engineering teams
2. AI infrastructure & orchestration (proxy, lambdas, streaming)
   • Be able to follow existing architecture and design correct approaches for our current system
   • Work comfortably with Python services, Fast API, REST APIs, streaming, and deployment constraints
   • Collaborate on infra topics without being pure MLOps
3. AI quality, evaluation & alignment with business logic (refinement, support, bug fixes, continuous improvements)
   • Design and run evaluation pipelines for AI features, including continuous monitoring and quality upgrades
   • Implement custom, business‑logic‑driven checks for LLM outputs
   • Perform RAG / LLM evaluation (accuracy, consistency, hallucination controls)
   • Write and maintain validation logic and guardrail code, not only “prompt tweaks”
4. AI governance & EU AI Act compliance
   • Work comfortably under EU AI Act constraints and internal governance
   • Document model usage, data flows, and risks for each AI feature
   • Define guardrails together with Product/Security teams
   • Participate in training, documentation, and audits – owning building models and safe/secure AI operations
5. Observability, monitoring & production robustness of AI features
   • Read and interpret logs/metrics for AI components
   • Specify and interpret monitoring for AI flows (latency, error rates, drift signals, usage patterns)
   • Understand versioning and backwards compatibility of AI outputs and contracts with other services
6. Cross‑team collaboration & stakeholder work
   • Communicate clearly with Product, Engineering, Security, and Business teams
   • Translate DS/ML decisions into clear Jira tickets and Confluence docs
   • Walk customers/CS through AI issues, limitations, and upcoming improvements when needed
   • Lead or co‑lead refinements, design reviews, and post‑mortems for AI features
7. Research & exploration (models, methods, vendors)
   • Research and deliver focused PoCs/MVPs based on product/tech roadmap
   • Stay up‑to‑date with LLM / RAG / evaluation research (papers, blogs, open‑source) and translate relevant ideas into shippable solutions
   • Justify why a specific model / architecture / vendor is chosen for our procurement use cases
   • Maintain a research backlog, prioritise by product impact and feasibility, and document decisions/learnings
8. Multi‑agent systems & advanced reasoning (agentic frameworks and tooling):
   • Multi‑Agent Systems & Advanced Reasoning – Hands‑on experience with agentic frameworks (e.g., LangGraph, CrewAI, AutoGen) and advanced reasoning techniques (Chain‑of‑Thought, ReAct, Plan‑and‑Execute) to structure complex AI workflows
   • Integration & Tool Management – Ability to define safe APIs, connect agents to external systems, and implement memory/state management (short‑term and vector‑based) along with RAG pipelines
   • Programming & Infrastructure Expertise – Expert Python skills and deployment experience using Docker, Kubernetes, and cloud platforms (AWS, GCP, Azure) with monitoring via tools
   • Evaluation, Safety & Ethics – Designing benchmarks to measure agent performance, ensuring reliability and cost‑efficiency, while implementing guardrails to prevent prompt injection and maintain ethical, legal compliance
   • Strategic & Soft Skills – Translating business goals into agent workflows, critically evaluating AI outputs for errors or bias, and communicating complex AI reasoning effectively to stakeholders