Job Title

We’re building coSow — a new intelligence layer sitting upstream of enterprise software implementation spend.Enterprise software implementations represent hundreds of billions in annual global spend. Yet scoping, pricing, and vendor selection remain manual, inconsistent, and opaque.coSow replaces:Manual discovery callsSubjective scopingWeeks-long proposal cyclesWith:Structured scoping engineRetrieval-augmented estimationDeterministic rule-based modelingAI-assisted SOW generationA competitive marketplace for implementation partnersRisk scoring and benchmarking over timeThis is not an “AI wrapper.” This is a hybrid system combining:Structured domain modelingRAG pipelinesLLM reasoningDeterministic rule enginesMulti-tenant SaaS architectureMarketplace mechanicsThe RoleWe are looking for a senior engineer who thinks like a systems architect.You will design and ship:RAG workflows over historical implementation dataPrompt pipelines grounded in structured ERP domain dataRule-based estimators that parse and validate LLM outputMulti-tenant RBAC architectureMarketplace bidding logicSecure production deploymentStack:React + TypeScriptSupabase (Postgres, RLS, Edge Functions)LLM APIsVector search (pgvector or Pinecone)StripeSecure SaaS infrastructureWhat Makes This InterestingReal domain complexity (ERP implementations are non-trivial)Deterministic + probabilistic hybrid architectureEnterprise-grade trust requirementsMarketplace design problemsOpportunity to define the intelligence layer of a massive services ecosystemWho You Are5+ years experienceStrong in system designComfortable with ambiguityHave built or deeply understand RAG systemsUnderstand tradeoffs between LLM outputs and deterministic constraintsCare about production quality, not just demosCompensationCompetitive USD contract compensation (6 months)Optional equity discussion for long-term fitPotential transition into founding engineer roleIf you want to work on something ambitious at the intersection of AI, enterprise SaaS, and marketplace infrastructure - this is that opportunity.Share:GitHubArchitecture you’re proud ofYour view on LLM + rule-engine hybrid systems