AI Systems Engineer — Audit-Grade Investment Research Platform (Indian Equities) We are building an autonomous, institutional-grade “Investment Committee” system for Indian equities (NSE/BSE). This is NOT a trading bot and NOT a dashboard project. It is a backend-first, audit-grade research platform designed to perform forensic-level analysis of company filings, fundamentals, governance, and supply-chain signals. The system must operate with ZERO manual data handling and full traceability. ________________________________________ Core Objective Design and implement a fully automated, multi-agent research platform that: • Discovers, ingests, validates, and parses official exchange filings • Enforces evidence-based decision logic in backend code • Maintains cryptographically verifiable audit trails • Supports human approval ONLY for final buy/sell decisions ________________________________________ Key Technical Requirements (Non-Negotiable) • ZERO manual operations (no uploads, no extraction, no human data handling) • Logic-gated Supervisor with hardcoded BUY rules in Python (not prompts) • Strict evidence hierarchy (Concall = supporting only, never BUY trigger) • Secure fallback with domain allow-lists, SHA256 checks, timestamp validation • ISIN-first identity mapping with versioned, hashed document storage ________________________________________ Expected Technology Stack • Python (advanced backend) • LangGraph (stateful multi-agent systems) • Pydantic / schema enforcement • LlamaParse or equivalent document parsers • PostgreSQL • Vector databases (Pinecone / Chroma / FAISS) • Indian market data APIs (NSE/BSE, Zerodha, etc.) ________________________________________ Professional Standards This is an institutional-grade research system designed to run “lights-out.” Any proposal involving manual data handling or scraping will not be considered. ________________________________________ Candidate Profile We are seeking a Backend Systems Architect or Quantitative Data Engineer with experience in financial data pipelines, compliance-grade logging, and production AI orchestration. If your background is primarily UI dashboards, scraping, or prompt engineering, this project is unlikely to be a good fit. ________________________________________ Evaluation Shortlisted candidates may be asked to complete a short technical PoC demonstrating automated PDF ingestion, hashing, and validation. ________________________________________ How to Apply Please include: 1. Relevant past projects 2. High-level technical approach 3. Fixed-price quote 4. Estimated timeline (weeks)