About the roleYou'll take on a product area within core Kikoff as its data lead, working day to day with the product, engineering, design, and lifecycle marketing leads for that area, and you'll sit in the Kikoff-wide conversations on roadmap and objectives.
Two things we're asking of this hire beyond the product area. First, help set technical direction and best practices for data science across Kikoff: how we do experimentation, how we evaluate AI products, how we review each other's work. Second, help define how we work as AI agents become a core part of the analysis loop, from exploration to pipelines to experiment readouts. We're actively rebuilding our workflow around this and want someone who has opinions.
What you'll do- Lead the data work for a core Kikoff product area: set the questions worth answering, build the evidence, and drive what happens next. Sometimes the right call is not to act on a finding, and you'll make that case too.
- Define and maintain the measurement system for your area across the whole customer journey (activation, engagement, credit outcomes, retention, revenue, unit economics), and contribute to the Kikoff-wide measurement framework alongside the other data scientists on the team. Where acquisition intersects with what you own, you'll work it jointly with Marketing DS rather than around them.
- Own product experimentation for your area: design, guardrails, analysis, and the recommendation on rollouts, including the cases where a holdout isn't clean or the effect you care about (a customer's score) moves on its own schedule.
- For AI product surfaces, own evaluation: decide what good means in checkable terms, build and validate automated scorers against human judgment, and turn what you find in real conversations into regression tests so the product can't quietly get worse. Keep the loop between error analysis and the eval set closed.
- Build and evaluate models where they're the right tool: proof-of-concept and challenger models, offline evaluation, threshold decisions, and production monitoring with engineering. Production model lifecycle sits with engineering today; how we divide that work is still evolving and you'll have a voice in it.
- Partner with product, engineering, design, and lifecycle marketing leads on roadmap and objectives: which bets, what a win looks like, and what we'd need to see to stop.
- Raise the bar for the people around you: review work, onboard new teammates, and take on an intern or early-career data scientist when the timing fits.
Minimum qualifications- Experience partnering with product, engineering, and marketing peers across the whole arc of the work: strategy, goal setting, approach, and execution, not just the analysis at the end.
- A track record of defining metrics from scratch and getting a team to run on them, including for products where success was hard to pin down.
- Designed and ran experimentation programs, including changes where clean randomization wasn't available. Comfortable with quasi-experimental and causal inference methods, and clear about their limits.
- Hands-on with production-quality SQL and Python. You build pipelines, analyses, and models yourself.
- Experience building or working closely with models that drive decisions in a product, in any domain: ranking, fraud, forecasting, personalization, underwriting, detection, LLM applications. We care about the judgment, not the vertical.
- AI tools are a core part of your daily analytical work and you can show how they changed the speed and quality of what you ship.
- You drive decisions with data in front of senior audiences, including when the data doesn't support the plan.
Preferred qualifications- Built or ran an evaluation program for an LLM-based product: judge design, validation against human labels, test-case construction from real failures.
- Consumer fintech experience, especially products that expand access for un- and under-banked customers.
- Built an experimentation or causal inference practice in an org that didn't have one.
- Have taken a model from proof of concept to production, or shipped test and challenger models that changed a product decision.
- Have mentored, onboarded, or managed the work of other data scientists.
Base Range
$226,000-$254,000 USD
