Data Science Lead

Elliptic is the leader in digital asset decisioning, with the most comprehensive platform for efficiently extracting crypto data and intelligence across blockchains with the greatest accuracy. Founded in 2013, Elliptic is headquartered in London with offices in New York, Washington D.C., UAE, Singapore and Tokyo. Our platform's unrivalled uptime, scalability, depth and breadth of data and intelligence means exacting organizations choose Elliptic for their compliance, risk management, intelligence operations and blockchain infrastructure needs.

We are looking for a leader to lead data science inside Elliptic's Intelligence team, alongside Intelligence Collection, Research and Investigations, and Professional Services. This is deliberately a player-coach role with a prospect of scaling a team. You will lead a small team of data scientists and own the team's direction, performance and growth. You will own the evidence for the models your team builds, and represent them at Elliptic's model risk governance forum.

A small team working on an unsolved problem needs someone who intrinsically understands the domain, not a layer of management above it. You will set what this team works on, hire the next people into it, and be the person whose technical judgement the team trusts because you can do the work yourself.

• Lead and grow the team. Set objectives with your data scientists in the context of company strategy, own their performance and development, and be the manager who has the difficult conversation rather than deferring it. Level the team against Elliptic's Intelligence career framework and build real development paths.
• Hire the next data scientists. Own the hiring process end to end with support: levels, job descriptions, hiring plan, interview panel and bar, and the hiring decision. Team growth to seven to nine people over time depends on you getting this right early.
• Build the collection for the future of compliance. Take it from a directive to a specification to delivered data: define what good looks like, sequence the work, set the quality bar, and be honest about what the dataset does and does not support.
• Own model governance for your team's models. Keep the model inventory current and correctly tiered, make sure validation and testing evidence exists at the standard each tier requires, monitor for drift and act on threshold breaches, and submit material changes and production deployments for review with the testing artefacts attached. Own the remediation of issues raised against your models, to agreed deadlines.
• Lead Elliptic's data science work. Frame the research question, run the first experiments, and form a defensible view of what Elliptic needs to be able to do and by when. Bring that view to Intelligence leadership, the Chief Scientist and Product as a recommendation rather than as an open question.
• Stay technically hands on. Write code, review your team's work at the level of method rather than output, and make the method calls yourself when they matter.
• Own the team's existing commitments. The team's current work on clustering, attribution, heuristics and labelling continues to serve the platform. You own its reliability and its trade offs, and you decide what to stop.
• Work across functions. Partner with Intelligence Collection, Collection Engineering, Research and Investigations, the Chief Scientist, Product and Engineering, so that data science work lands as platform capability rather than as analysis nobody uses.
• Set the AI working standard for the team. Decide how your team uses AI in its engineering and research workflow, what must be verified and how, and hold the team to it.

• The collection plan delivered at a scale that supports real model or heuristic work, with a documented path to maintaining it.
• An agreed Elliptic position on the evolution of compliance systems, timed against expected industry demand, with the first capability work either underway or deliberately deferred with stated reasons.
• At least one data science hire made, and the existing team levelled against the career framework with development plans in place.
• Model governance in good standing for everything your team owns: inventory current and correctly tiered, no overdue validations on the highest tiers, monitoring in place with acted-upon thresholds, and no open issues past their remediation deadline.
• Data science objectives for Intelligence agreed and delivering,…