Founding Machine Learning Engineer

About the role

ABOUT CENTRALIZE

Enterprise sales runs on relationships, and every tool built to manage them is a database from 2005. Reps lose seven-figure deals because they can’t see who actually matters inside an account. We’re building the system of intelligence for relationships, the R that CRMs never actually had.

Centralize is the relationship intelligence platform for enterprise revenue teams. Companies like Webflow, Decagon, ElevenLabs, Brex, Cognition, and LangChain use us to close their largest deals. We’ve grown 5x since last year, and customers are pulling us forward faster than we can ship.

We’ve raised a Series A led by NEA, bringing our total funding to $19M alongside Salesforce Ventures, Y Combinator, and operators including Stewart Butterfield (Co-Founder & CEO, Slack), Cal Henderson (Co-Founder & CTO, Slack), and sales leaders from Figma, Box, Dropbox, Anthropic, and Notion.

Centralize was founded by Rachit Kataria, a founding engineer on Facebook Shops who helped scale it to 250M MAUs, and Will Wang, who led the launch of Slack Huddles, the fastest-growing product in Slack’s history.

THE ROLE

We’re hiring our first ML engineer to lead ML and applied AI across Centralize, starting with our relationship graph: who knows who, who trusts who, and how it shifts deal by deal. You’ll report to Will Wang, our co-founder and CTO, and build a new research arm from the ground floor. You’ll set the research agenda, own entire systems end to end rather than one slice of a ranking model, and grow into leading the ML team as it forms around you. Your models learn from every customer deal and ship within weeks, and we’ll publish what we find, with our first white papers targeted for Q1 2027.

This is not a traditional post-training research role. The work is recommender systems, knowledge graphs, and learning from past deals, applied in production. Classical ML, recsys, ranking, and graph backgrounds are all good fits, and you’ll flex into applied AI when the problem calls for it.

We have two openings:

- Senior: you set the agenda, own the first systems, and grow into leading the team. Starts before the end of 2026.

- Mid-level: you go deep on large parts of the system alongside the senior hire and grow fast. Starts by the end of 2026 or in January 2027.

WHAT YOU’LL DO

- Set the research agenda for Centralize’s trust graph.

- Design and build the layered knowledge graph and the recommendation systems on top of it: embeddings, retrieval, ranking, and link prediction that improve with every deal, signal, and outcome.

- Build next best actions: learn from past deals and outcomes to recommend who a rep should reach out to, why, and how.

- Drive research in production. Take ideas from first experiment to shipped feature in front of customers, and measure the impact in real deals.

- Own system design and model selection: decide when a third-party model or API is the right call and when to build in-house, and build the data, evaluation, and inference foundation that keeps it fast and reliable at enterprise scale.