Full‑Stack AI Engineer and Project Manager (Site + Chatbot + Questionnaire + Auto Content Engine)

Summary

We’re building a B2B Planner that helps campground developers and hospitality pros discover suppliers/products, get consultative answers via a GPT‑powered chatbot, generate a decision‑grade report from a questionnaire, and (later) run automated content ops.

You’ll implement and manage the whole build across three phases. The datasets (suppliers + products and supplementary data) and prior scope documents are ready for you.

## What you’ll own ##
### Phase 1 — Website + Chatbot (ship first): ###
- Public site (clean, outdoor‑inspired look & feel; exact pages are up to you).
- Embedded chatbot using GPT‑5 (or equivalent) with RAG over our supplier & product data + post‑purchase knowledge base.
- Basic admin tools to inspect retrieval/citations and tweak prompts.

### Phase 2 — Questionnaire + Report: ###
- Pro‑mode questionnaire that maps answers to supplier/product matches, budgets, legal pointers, timelines, features, and positioning.
- Map step (optional): user either draws their parcel live on an embedded Google Map or pastes a map image/screenshot; we compute/confirm area and generate a structure plan.
- Quick mode: user provides only the area (from drawing or image) + a few toggles to get an instant plan.
- A report (HTML + PDF) used by both full questionnaire and quick mode.

### Phase 3 — Auto Content Engine (ACE): ###
- Must be possible to run entirely in n8n (no custom schedulers required): topic scraping/ingestion, brand‑voice angle generation, scheduling to blog + Instagram + Facebook + X, and a feedback loop that adapts the plan. You’ll wire n8n workflows plus minimal glue code as needed.



You’ll also handle basic analytics, QA passes, and lightweight documentation so the system is easy to operate.

## Inputs you’ll get ##
- Data: suppliers_with_ids.csv, product_specifications.csv, plus a curated post‑purchase knowledge base and other supplementary datasets.
- Design reference: Sleepless Media’s Outdoor Industry style (clean, modern, outdoor‑focused). Exact site structure is up to you.

## Recommended stack (flexible — propose alternatives if you prefer) ##
- Frontend: Next.js + React (TypeScript), Tailwind for speed.
- Backend: Next API routes or Node/Express.
- RAG: Postgres + pgvector (or any solid vector DB you like).
- PDF: Server‑side HTML→PDF (Playwright/Puppeteer).
- Map: Google Maps JavaScript API with Drawing tools (live polygon draw) and image‑paste support (user provides the area from the screenshot overlay).
- ACE: n8n for scraping, processing, generation, scheduling, and reporting (first‑class requirement).
- Dev tooling: AI‑accelerated workflow strongly encouraged — Replit, Cursor, Copilot, Codeium, etc. Quick turnaround is valued over heavy custom frameworks.

If you can deliver faster or more maintainably with a different stack, say so. The tech is recommended, not mandated.

## Must‑have skills: ##
- Shipping production web apps end‑to‑end as a solo or lead engineer/PM.
- LLM RAG (chunking, hybrid retrieval, citations, prompt/guardrails).
- Google Maps embedding and polygon drawing UX; ability to accept a pasted map image and translate the user‑provided area into the plan.
- n8n expertise for end‑to‑end content workflows (scrape → cluster/score → generate angles → schedule → measure → adapt).
- HTML→PDF; basic analytics; GDPR‑aware data handling.

## Nice to have: ##
- Tailwind craftsmanship; accessibility basics (WCAG AA); CMS experience; social platform APIs; CRM/webhook integrations.

## Milestones (you quote timelines & price) ##
- Phase 1: Site + Chatbot (public site online with embedded chat + admin playground).
- Phase 2: Questionnaire + Report (includes Map step and Quick mode; unified 10‑section report).
- Phase 3: ACE (n8n) (topic intake, generation, scheduling, feedback loop + simple editorial console).

I’ll review and approve each milestone as Supervisor before we proceed to the next.

## How to apply ##

Please send:
- 2–3 links to relevant builds (LLM/RAG or complex full‑stack).
- A short note on how you’d implement ACE entirely in n8n (one paragraph or a simple diagram).
- Your preferred stack (if different) and how AI dev tools will speed delivery.
- Your high level milestone estimate for phase 1 (time + price) and earliest start date. Feel free to make any assumptions on data and site structure.

That’s it. If it’s a fit, we’ll jump to a quick call and get going.