Do you want to lead a data science team on a problem the industry has not solved yet, while still writing code yourself? Are you excited about building the data foundation for how compliance works when the customer is an autonomous agent rather than a person? Are you looking for a values led company where a small team is trusted with a question that matters commercially in months, not years?The impact you will have:You will 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. At first you will lead a small team of data scientists and you are expected to own the team's direction, performance and growth. You will own the evidence for the models your team builds, and you will represent them at Elliptic's model risk governance forum. This is not a compliance tax bolted onto the technical work: it is the reason institutions trust our outputs.
The reason we are hiring a lead as we add more data scientists is that 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.
What you will do:- 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.
What you will achieve in the first 6 months:- 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 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, with other functions able to state what they can expect from the team and when.
- A visible improvement in the reliability or throughput of the team's core output, attributable to a change you made.
- Technical credibility intact: the team takes your method judgements seriously because you can do the work.
You will be a great fit here if you:- Want to lead a team without leaving the technical work behind, and can say honestly where you draw that line.
- Are comfortable being accountable for a question nobody has answered yet, including for saying what you do not know.
- Treat management as the work rather than as overhead: objectives, feedback, development and difficult conversations included.
- Treat governance as part of building good models rather than as paperwork that follows them, and would rather surface a problem with your own model than have someone else find it.
- Are pragmatic about trade offs between rigour and speed, and can explain the trade off you chose.
- Make other teams more effective rather than treating them as consumers of your team's output.
- Are curious about how the company operates beyond your own function.
- Care about the mission of using intelligence to fight financial crime and protect the global financial system.
Our ideal candidate has:- AI fluency, which is essential for this role. You must be able to demonstrate how you apply AI tools and approaches within a data science workflow: writing and debugging code, exploring unfamiliar data, accelerating pipeline and model work, and automating repetitive tasks. You must also be able to show how you critically evaluate the output, where you would not rely on it, and how you set verification expectations for a team rather than only for yourself. This is assessed in a dedicated stage of the interview process, working with your own tooling.
- Demonstrated experience managing data scientists or machine learning engineers: setting objectives, owning performance including underperformance, and developing people. You can point to someone whose career changed because of how you managed them.
- Experience hiring into a technical team, including defining the bar rather than only sitting on panels.
- Deep, hands on data science and machine learning capability that is current. Strong Python and SQL, and the ability to interrogate large behavioural or transactional datasets yourself and defend your method under challenge.
- A track record of taking an ambiguous question to a delivered dataset, model or capability, including deciding what was good enough.
- Experience with a modern data stack comparable to ours: a cloud data lake, Spark or Databricks, and AWS.
- Clear communication with technical and non technical stakeholders, and the ability to cascade direction to a team so that each person knows why their work matters.
- Based in Washington, D.C. or willing to relocate there.
- US citizenship
Bonus Points for:- Work on agentic systems, LLM agents or autonomous transaction flows, whether in compliance, payments, fraud or infrastructure.
- Experience building a large dataset or collection from nothing, including the unglamorous parts: schema decisions, quality checks, and maintenance.
- Blockchain or on chain data experience: clustering, attribution, heuristics, anomaly detection, mempool analysis, or research into privacy preserving systems.
- Experience in payments, fraud or risk data science, where the cost of a false negative is commercial rather than academic.
- Having grown a team from a few people to a full department, and being able to describe what broke on the way.
- A PhD or equivalent research training, or a record of published applied research.
- Experience in a regulated environment or with model risk management frameworks, whether in financial services, or as a vendor to it, or in another domain with equivalent assurance requirements.
Benefits:> How we work:
- Hybrid working: The option to work from almost anywhere for up to 90 days per year.
- Remote Work Budget: $650 budget to set up your home office space.
> Learning & Development:
- L&D Budget: $1,000 annual Learning & Development budget to use on anything (agreed with your manager) that contributes to your growth and development.
> Vacation / Leave:
- Holidays: 25 days of annual leave + 8 US Public Holidays.
- Birthday Leave: An extra day off for your birthday.
- Enhanced Parental Leave: We provide eligible employees, regardless of gender or whether they become a parent by birth or adoption, 16 weeks fully-paid leave.
> Benefits:
- Healthcare: Comprehensive medical, dental, and vision coverage through a range of providers (including Tufts, Kaiser, Aetna, UHC, and Blue Shield of CA) with generous premium contributions for you and your dependents.
- 401k: Company match included.
- Mental Health: Full access to Spill mental health support.
