Alpha Signal, LLC Backend Engineer Remote · Full time Company website

Apply for Backend Engineer

AlphaSignal is building the intelligence layer for AI - a real-time, machine-readable map of everything happening in the field, linked into a single knowledge graph. We run the first real-time AI ranking system and the first autonomous editorial system in the world. Our newsletter reaches 300,000+ subscribers, making it the most-read independent AI publication on the internet.

THE ROLE

You'll own the pipeline behind the platform: scrapers, an LLM enrichment and editorial layer, a MongoDB data layer, a ranking system that decides what matters, and the foundations of our knowledge graph.

You're responsible for the code feeding a product used by over a million people a month. When throughput drops or something breaks, you diagnose it, ship the fix, and verify the recovery.

We're a team of 10. You see a problem, you fix it.

────────────────────────────────────────

WHAT WE'RE LOOKING FOR

• 3+ years of backend engineering with production ownership of a non-trivial system - you got paged when it broke.

• Python at a professional level - backend services, ML models, algorithms, and production-grade diagnostic scripting.

• Deep LLM fluency across providers (Claude, GPT, Gemini) - shipped production features, fluent with prompting, structured output, cost optimization, and failure modes. You treat prompts as engineering artifacts, not vibes.

• Database depth - MongoDB, Postgres at scale, or comparable; indexes, query plans, atomicity, replication.

• ML / vector search - recommendation systems, knowledge graphs, entity linking, and the full vector DB landscape (Pinecone, Weaviate, Qdrant, Milvus, pgvector, Chroma, FAISS, plus Atlas/Elasticsearch/OpenSearch).

• Designed systems for large, growing traffic - caching, batching, async, and index tuning before things fall over.

• AWS / infrastructure - comfortable deploying and operating services on EC2 and the broader Amazon suite.

• Fluent with modern AI dev tooling (Cursor and similar).

• Solid engineering practices: observability, monitoring, ticketing, and keeping a codebase healthy as it scales. Ideally worked on a team of 5+ developers.

────────────────────────────────────────

BONUS POINTS

• Experience at a high-traffic content platform (Medium, Substack, Reddit etc)

• Familiarity with arXiv, GitHub, Hugging Face, or X APIs.

• Content systems: feeds, aggregators, recommenders.

• Scraping at scale, deeper prompt engineering, or early-stage startup experience.

────────────────────────────────────────

NOT A FIT IF…

You need a big team to build anything, prefer to specify rather than build, treat LLM calls as a black box, or are allergic to Python or MongoDB.

────────────────────────────────────────

TECH STACK

Python 3.11+, MongoDB Atlas (incl. vector search), Pinecone / Weaviate / Qdrant / Milvus / pgvector / Chroma / FAISS, LLM APIs (Claude, GPT, Gemini), AWS (EC2 + broader suite), Firecrawl, Replit, BeautifulSoup, arXiv / GitHub / Hugging Face / X, Cursor, Sentry, cron-scheduled services.

Non-negotiables going in: Python, LLM fluency, database depth, and AWS/infra comfort.

────────────────────────────────────────

WHAT YOU'LL GET

• Full ownership of a critical, high-impact system.

• Direct work with the founder.

• Real, visible impact in production — often within hours.

• Competitive compensation.

• Remote, async-first, low-meeting culture.

────────────────────────────────────────

HOW TO APPLY

Send a short note, a link or two to systems you've owned, and one debugging or design story where you got something interesting right - or wrong, and what you learned. We care about how you think, not boxes checked.