Who We Are & Project Scope We're a US-based B2B data management consulting firm (MDM, data governance, analytics, AI readiness) looking for an experienced full-stack developer to build a lightweight internal outbound prospecting tool from scratch. The system will run 75–100 highly personalized outbound emails per week (15–20 per business day), operated daily by an offshore operator, with a human approval step before each send. Quality over volume is key — replies are the success metric. A full technical spec is ready to share with serious candidates after an initial message. What You'll Build A small web app + background worker deployed on a VPS that lets an offshore operator run the daily workflow end-to-end. Core components include: - Lead management with deduplication and email verification - AI-powered company research engine generating sourced briefs (Claude or OpenAI) - Human-approval review queue before every send — no email goes out unreviewed - Direct Gmail API sending across 1–2 mailboxes, paced randomly across business hours, hard cap of 15/day per mailbox - Reply capture, LLM classification, and automated suppression/bounce handling - Follow-up sequencing of 3–5 touches over 2–4 weeks, with branching logic (OOO, referrals, positive replies) - Monthly AI optimization report for qualitative learning Tech -Python/FastAPI or Node/Next.js + Postgres (open to your recommendation). -Integrations: Gmail API (OAuth 2.0), Anthropic Claude or OpenAI, web search API, pay-as-you-go email verification. -The total recurring infrastructure cost target is under $100/month. What We're Looking For - Production web app experience with background job processing - Comfortable with LLM APIs and real-world prompt design - Strong on compliance details: idempotent queues, suppression logic, CAN-SPAM - Clear communicator who flags tradeoffs early Project Terms - Work-for-hire; all IP transfers to us - Fixed-price bids preferred, structured around 4 defined milestones - Target delivery: 4–6 weeks - Direct access to decision-makers throughout To apply Please describe a similar system you've built, your preferred stack and why, and any questions you have about the project. Candidates who ask specific technical questions are given priority.