Senior Full-Stack Engineer – LLM / NLP Focus

Why this role matters

AI search is reshaping how buyers discover software. xfunnel brings transparency to that black box - monitoring how engines like ChatGPT, Gemini, and Perplexity cite brands, then turning those insights into revenue-facing actions.

xfunnel.ai monitors that shift in real-time and turns it into revenue signals for marketers. To keep pace, we need a senior engineer who can own the full stack - while wrangling LLM pipelines that scale.

What you'll do

• 40 % – Ship product features end-to-end
  Next.js, TypeScript, Node, Supabase
• 30 % – Build LLM/NLP services
  Prototype, fine-tune, and serve models with different LLMs then harden them with tests and observability hooks.
• 15 % – LLMOps & scalability
  Containerise models, set up retrieval-augmented generation (RAG) and evaluation loops so features stay fast and cheap in prod.
• 10 % – Data & analytics plumbing
  Create pipelines that ingest data and user events, transform them into embeddings, and expose insights via dashboards.
• 5 % – Mentoring & code quality
  Review PRs, level-up junior engineers, and guard our architecture's simplicity as we scale.

You'll thrive here if you…

• Bring 5-8 yrs professional dev experience, with 3+ shipping modern full-stack web apps (React/Next.js + Node/Go/Python).
• Have hands-on NLP/LLM chops: fine-tuned a Transformer, built a agent, or shipped a RAG feature to prod.
• Treat CI/CD, automated tests, and "you build it, you run it" as non-negotiable.
• Love small-team ambiguity—one hour coding, next hour debugging prod, then jumping into a customer call.
• Write crisp, async-friendly prose (PRDs, RFCs, Looms) because half our team sleeps while you code.

Bonus points

• Previous work on search/SEO analytics, marketing funnels, or ranking systems.
• Experience with vector databases (Weaviate, PGVector) and structured logging for LLM observability.
• Familiarity with privacy-preserving ML or reinforcement-learning-from-human-feedback.

What you'll get

• Green-field canvas to architect the AI layer of a category-defining product.
• Founders who ship code and remove blockers, not create them.
• Competitive salary + meaningful equity + quarterly offsites (NYC / TLV / Madrid).

Interview process

• 30-min fit call with a founder.
• 90-min live coding & architecture (bring a past system to dissect).
• Async take-home (≤3 hrs) – build a tiny RAG or UI slice; we pay for your time.
• Panel chat with team → culture & roadmap Q&A.
• Offer – reference check, comp chat, start date.