Forward‑Deployed AI Engineer (LLM/ML)

Silent Eight is a leading RegTech firm developing AI-based products to combat financial crimes such as money laundering, terrorism financing, and systemic corruption. They work with large international financial institutions including Standard Chartered Bank and HSBC.

• A track record of delivery: 2–3 examples where AI/ML solutions materially improved process KPIs.
• Ability to break down fuzzy problems and explain risks to non-technical stakeholders.
• Practical experience with LLMs and ML: RAG, prompting, tool-calling; classification, ranking, deduplication; fundamentals of evaluation and experimentation.
• Proficiency in Python and SQL sufficient to build and maintain prototypes.
• Fluency in Polish and English for user conversations and concise write-ups.

Nice to have:

• Experience with investigations, trust & safety, fraud, risk, audit, or other complex decision processes.
• Knowledge of graphs/entity resolution, link analysis, pattern-of-life.
• Light engineering skills: FastAPI, Docker; Kubernetes, CI/CD, Terraform not required.

• Talk to users, shadow their workflows, capture the as‑is → goals & constraints.
• Define hypotheses & KPIs (precision/recall, FPR, TAT, coverage, cost/decision) and turn them into experiment plans.
• Design decision flows that mix LLMs, retrieval/RAG, classical ML, and lightweight rules; ensure explainability and auditability.
• Build quick prototypes (notebook → lightweight service/API) and measure their impact on real data.
• Create evaluation sets and scoring rubrics (offline + side‑by‑side + sanity checks + guardrails).
• Present findings & recommendations directly to decision‑makers; propose rollout (pilot → production‑lite → scale).
• Lead innovation processes across the company; test, promote solutions and mentor others with new AI technologies.