Senior AI Product Engineer - Build LLM-Powered Competitive Intelligence Web App

The project We're starting a new AI product from scratch and we need a senior AI engineer to build the entire thing: frontend, backend, database, and an AI/LLM reporting layer. Nothing is built yet. You'd be architecting and shipping the full platform end to end. This is a real product with a clear vision, not an experiment. To be clear up front: this is not a fit for juniors. We want people who have shipped production AI products and can own architecture decisions without hand-holding. What we're building A competitive intelligence platform for the online gaming and betting market. The app monitors how casino and betting games are ranked across multiple popular operator websites and tracks how those rankings move over time. The system needs to scrape these sites on a schedule, store a growing time-series history of every game's position, category, and provider, and present all of it as live rankings, trends, and per-operator breakdowns in a clean web app. The core differentiator, and the part we care most about, is the AI layer. The AI insight engine We want an LLM reporting service that reads the change between snapshots over time and writes a short, plain-English intelligence brief that a non-technical operator can act on. Instead of making users read tables, the system should detect what changed day over day and explain it in clear human language. So you'd design and build the full pipeline: compute snapshot diffs, structure that change data, feed it to the LLM with the right prompts and guardrails, and generate clean daily and weekly briefs that get stored and shown in the app. Output quality, accuracy against the source data, and cost control on the LLM calls all matter. We want this to read like a real analyst wrote it, not a generic AI summary. Full scope (greenfield build) Frontend: fast, responsive web app for rankings, trends, and per-operator views Backend: scraping layer plus scheduled jobs, running reliably at scale Database: time-series data model for snapshots over time AI/LLM reporting service: the insight engine described above (daily + weekly briefs) Deployment, monitoring, and scheduling for scrapers and report jobs Required tech stack Node, Next.js, React, Astro, Tailwind, Supabase, PostgreSQL, Vercel, Hetzner, AWS. LLM integration via OpenAI and/or Anthropic APIs. You should be comfortable choosing the right tool for each layer and defending the choice. Who we want Experienced, AI-proficient and senior engineers who use AI tools for rapid development: Cursor, Claude Code, Codex, and similar. We care about shipping speed and quality. Strong candidates have shipped real LLM-powered products, understand prompt engineering and evaluation, and can reason about token cost, latency, and accuracy. Clean code and sound architecture judgment are expected. Nice to have Scraping at scale and anti-bot handling, embeddings and semantic search, time-series data modeling, and prior work in analytics or business intelligence products.