We need to build an AI Lead Agent for our travel-tech platform (PerlaHub). Tech stack required: • Mastra (TypeScript/Node) → Agent runtime • Python (FastAPI) → Tools for DB/API/Email • PostgreSQL • REST APIs • OpenAI • Optional: LangChain (only as library, not agent runtime) Project: Build a fully autonomous Lead Agent that creates demo credentials for new commercial leads. The agent must: • Interact conversationally (chat UI) • Validate required fields (name, market, distribution, product feed, email, username) • Use an existing Demo Client (never create new clients) • Create credentials in TEST mode • Write to PostgreSQL tables (smartperlahub.leads, partners.ClientCredential, etc.) • Call our PerlaHub API to create/configure credentials • Generate username/password/URL • Prepare and send email with credentials • Update lead status (“In progress → Keys sent → Benchmark”) • Log all actions (input, output, decisions, errors) Deliverables: 1. Lead Agent defined in Mastra (TS) 2. Python microservice for tools (DB, API, Email) 3. Integration with Latitude (evals, datasets, observability) 4. Documentation + test cases Looking for: A developer experienced in TS + Python, building real LLM agents (not basic chatbots), and able to connect to APIs and databases.