Senior AI Systems Engineer – Build Persistent Reasoning Engine (LLMs, Knowledge Graphs, Python)

Senior AI Systems Engineer – Persistent Reasoning Engine I'm looking for an experienced AI systems engineer to help build the first version of a persistent reasoning engine. This is not a chatbot, RAG application, or document summarization project. The objective is to build a system that ingests evolving public documents (SEC filings, annual reports, shareholder letters, earnings call transcripts, and news) and maintains an evolving enterprise knowledge model over time. Each new document should update the existing enterprise state rather than generate a new isolated summary. Think of it as maintaining continuity for an organization instead of summarizing individual documents. Version 1 Goals The first prototype should: • Ingest SEC filings (10-K, 10-Q, 8-K) • Ingest shareholder letters • Ingest earnings call transcripts • Parse PDF and HTML documents • Use LLMs to extract structured information • Store a persistent enterprise record • Compare current state with previous years • Track how enterprise state changes over time The output is not a stock recommendation. The output is a structured enterprise model that includes: • Doctrine • Capabilities • Active States • Active Obligations • Risks • Management Decisions • Causal Relationships • Enterprise Trajectory Preferred Technical Experience • Python • FastAPI • PostgreSQL • OpenAI and/or Anthropic APIs • Production LLM applications • NLP / information extraction • Knowledge graphs or graph-based reasoning • Vector databases are a plus What I'm Looking For I'm looking for someone who enjoys solving difficult architecture problems rather than simply integrating APIs. If you've built production AI systems involving persistent state, temporal reasoning, knowledge graphs, or agent workflows, I'd like to talk. This prototype is the first validation of a larger reasoning architecture that may eventually be applied in additional domains. First Deliverable The first milestone is not a polished application. It is a working prototype capable of ingesting Microsoft and Amazon annual reports, extracting structured enterprise objects, and maintaining an evolving enterprise state across multiple years. Please include in your proposal Please begin your proposal by answering one question: "What is the difference between a document summarization system and a persistent reasoning system that updates an evolving knowledge model over time?" Also briefly describe how you would architect this