Senior .NET Developer — Rule/Decision Engine (PostgreSQL)

Please read before applying. We're a small team and receive a high volume of applications — we can't screen everyone. Only apply if you genuinely have the expertise described below, can point to real experience with it, and are ready to be productive from day one. This role does not include on-the-job training. Generic proposals, or applications that don't address the required skills, will not be reviewed. We're building a decision engine for automated financial data processing — a .NET 8 backend that evaluates authored business rules against structured data, with a strong emphasis on correctness, auditability, and clean architecture. This is a from-scratch build of the core engine, not maintenance of an existing app. We're looking for a senior .NET engineer who's comfortable working close to the language: parsing, evaluation, and precise data modeling, not just CRUD. What you'll work on • A rule-evaluation runtime: evaluating authored rules against dynamic, schema-flexible input where field names aren't known until parse time • A small domain-specific language for authoring rules — lexer, parser, and mapping to an internal representation • Similarity-based matching using vector embeddings in PostgreSQL, alongside exact rule matching • A full audit trail for every decision — persistence, versioning, and effective-dating of rule sets • PostgreSQL schema, query performance, and data integrity, including pgvector Required • 5+ years professional C# / .NET (production systems, not tutorials) • Strong hands-on .NET 8 • Strong PostgreSQL — indexing, query tuning, transactional design; Npgsql and Dapper experience preferred over heavy-ORM reliance • Experience building complex business logic — rule engines, decision tables, expression evaluation, or interpreters/DSLs • Solid grasp of writing testable code (xUnit or similar) and integration testing against real databases • Clear written English Nice to have • Experience with pgvector or vector similarity search • Parser/tokenizer work (e.g. Superpower, ANTLR, or hand-rolled) • Familiarity with a .NET rules engine substrate (e.g. Microsoft.RulesEngine) • Frontend React — not mandatory, but a plus if you can also build a UI on top of the engine • Resilience patterns (Polly), structured logging (Serilog) • Background in fintech, accounting, ERP, or financial data systems • Cloud experience in a regulated environment Screening questions 1. Describe a system where you implemented complex business rules, an interpreter, or a decision engine. How did you structure the logic to keep it testable and maintainable? 2. Have you built or worked on a parser or DSL? What was the input, and how did you handle malformed or edge-case input? 3. How would you approach matching records where some cases resolve via exact rules and others need a similarity/fallback strategy with a confidence threshold? 4. Have you designed a system that keeps a full audit trail and supports versioned, effective-dated configuration or rules (i.e. "which rule set was active on date X")? How did you model it?