Causa Prima is a B2B payments network that connects buyers and suppliers so AI agents can automatically handle invoicing, dispute resolution, and payment-term negotiations.
You'll be core to our product — designing, building, and operating the autonomous agents that power Causa Prima. From invoice validation and dispute resolution to cross-company negotiation and cash optimization. What you'll do - Agent architecture — Design the multi-agent system: agent boundaries, event contracts, orchestration patterns, and the separation between LLM reasoning and deterministic enforcement. LLMs inform; rules enforce. - Document processing pipeline — Ingestion, classification, and structured extraction from financial documents (invoices, contracts, purchase orders) using LLM-based parsing and vision models. Sandboxed processing for untrusted documents. - Validation & anomaly detection — Business rule checks, three-way matching, pricing discrepancy detection, and anomaly detection across time. - Integration framework — Gmail, Google Drive, Slack, banking APIs, crypto wallets, and accounting platforms. Auth model, credential lifecycle, data normalization. - Knowledge graph — Neo4j context graph that captures implicit business knowledge — the “why” behind decisions. Entity extraction, GraphRAG patterns, provenance tracking. - Multi-LLM strategy — Per-agent model selection, structured output contracts via Pydantic schemas, evaluation framework, fallback strategy. - Compliance & security — Agents operate within a zero-trust model. You’ll design how agents verify independently against source data, how outbound content is reviewed, and how the system prevents cascading failures from prompt injection. What we're looking for - 3+ years experience with Python and/or TypeScript in production. - Strong systems design thinking — agent boundaries, failure modes, trust boundaries, contract evolution. - Strong interest and hands-on experience in LLM-powered systems, agent orchestration frameworks (LangGraph, LangChain, or similar), and document AI (LlamaIndex, LlamaParse, or similar). - Strong opinions about when to use an LLM vs. a rules engine vs. traditional ML — and the courage to defend them. - Experience designing data pipelines, integration architectures, or event-driven systems in production. - Security awareness — you think about what happens when untrusted input meets your pipeline. - Effective use of AI coding tools with strong review skills for LLM integration and auth code. - Interest in the finance domain — financial workflows, accounting processes, payment systems — is a strong plus.