Description
Summary Title: Senior AI/ML Engineer - Enterprise Knowledge Retrieval (RAG / LLM) for Stealth Startup Overview: We're an early-stage startup building an AI-powered knowledge layer for enterprises, a system that connects to a company's scattered data sources and lets people ask plain-language questions and get accurate, source-cited answers across everything. Think "enterprise brain": making an organization's institutional knowledge instantly findable and trustworthy. We're looking for a strong engineer to help build the core retrieval and reasoning engine. This starts as a contract role with strong potential to grow into a longer-term or equity-based relationship for the right person. What you'll work on (high level): Building retrieval-augmented generation (RAG) pipelines over messy, real-world enterprise data Connecting to and ingesting from multiple enterprise data sources (document stores, email, cloud drives, databases) Making answers accurate and traceable to their source - citations and verifiability are core, not optional Handling permissions, security, and multi-source data in an enterprise context Skills we're looking for: Strong experience with LLMs and RAG architectures (embeddings, vector databases, retrieval pipelines, orchestration frameworks like LangChain/LlamaIndex or custom) Building production AI applications, not just demos - accuracy, evaluation, reducing hallucination Data integration experience: ingesting/normalizing data from varied enterprise sources and formats Comfort with enterprise concerns: permissions, access control, data security/privacy Backend engineering (Python strongly preferred) and API/systems design Bonus: experience with cross-source / cross-entity data reconciliation (querying across separate systems as if unified) Ideal (not required): Familiarity with M&A, private equity, or corporate development contexts or enterprise environments where information is fragmented across many systems and teams Experience building tools used by non-technical business/enterprise users About working with us: Small, fast-moving founding team. You'd work directly with the technical co-founder. We value people who ship, care about accuracy and craft, and can operate with ambiguity. If this becomes a great fit, there's a path to something bigger. To apply: Share relevant RAG/LLM projects you've built (especially anything involving messy multi-source data or enterprise retrieval), and a note on your experience making AI outputs accurate and trustworthy.