GenAI Architect

Що робити

Scope and estimate (presales). Turn client briefs into structured requirements, assumptions, risks, and dependencies. Produce work breakdowns and estimates, propose team composition, and present side-by-side platform or approach options when the client needs to make a choice. Defend the numbers to stakeholders and adjust them as the scope clarifies.
Design the solution. Architect GenAI applications end to end: multi-agent orchestration, RAG and semantic-search flows, data and document connectors, evaluation harnesses, audit trails, and the surrounding web application. Choose models, platforms and patterns deliberately, and write the reference architecture and decision memos that capture why.
Be the technical voice with clients. Translate effectively between business stakeholders and engineers, reflecting the client's specific vocabulary and making complex trade-offs clear: such as choosing between a thin-slice POC and an enterprise data foundation, lightweight orchestration versus heavyweight, or build-now versus design-for-later.
Stay hands-on in delivery. Set up the technical foundations, prototype the high-risk parts, review code, and unblock the team. Build proofs of concept quickly to validate feasibility and value before scaling. Keep the delivered system true to the design — especially the parts that matter most for accuracy and traceability.
What we're looking for

Experience delivering into regulated or compliance-heavy domains — healthcare, life sciences, finance — where audit trails, data-handling constraints and traceability are first-class requirements.
Familiarity with observability & evaluation tooling and practices for LLM systems (accuracy baselines, eval harnesses,

Знайдено через trackr.help/jobs · Канал: @trackrhelp · Бот для персональних сповіщень: @trackrhelpBot