About the role
Own AI-native product features end to end in Quillstone's merchant and operations dashboard: talk to users, write the spec, build the React front end and the API, instrument it, and decide with data whether it stays. You will work in a small squad where engineers make product decisions.
Benefits & perks
Health Insurance
Private medical insurance (Bupa) for you and dependants.
Equity
Meaningful stock options, 4-year vest with 1-year cliff.
Flexible Work
Hybrid: 2 days a week in the Shoreditch office.
Paid Time Off
30 days holiday plus bank holidays.
Parental Leave
26 weeks fully paid parental leave.
Wellness / Stipend
£100 monthly wellness stipend.
- 01Build full-stack features in TypeScript, React, Next.js and Node.
- 02Integrate LLM capabilities with streaming UX and structured outputs.
- 03Run user interviews and turn findings into specs.
- 04Instrument features and run experiments to decide what to keep.
- 05Maintain offline evaluation suites for AI features in CI.
What we evaluate against.
Product Judgment & Prioritisation
Must have
- Has decided to cut or reshape a feature based on user or metric evidence.
Nice to have
- Writes their own specs.
AI & Technical Fluency
Must have
- Has shipped an LLM-powered feature to production users.
- Can describe a failure mode of that feature and how the UX handles it.
Nice to have
- Maintains evaluation or prompt tests in CI.
User & Workflow Insight
Must have
- Has run user interviews and changed a design because of them.
Nice to have
- B2B or finance user research experience.
Experimentation, Evaluation & Learning
Must have
- Has run an A/B test or evaluation and acted on a null or negative result.
Nice to have
- Uses product analytics tooling day to day.
Cross-Functional Product Delivery
Must have
- Has shipped full-stack features end to end with measurable outcomes.
Nice to have
- Resolved a design/engineering trade-off that shipped.
Team Structure
AI platform squad of 8 engineers working alongside disputes and reconciliation product teams.
Operating Rhythm
Two-week sprints, weekly customer calls, monthly model-risk review.
Communication
Written-first; design docs precede every customer POC.
Conflict Resolution
Disagreements settled on evaluation data, escalated to the CTO only on regulatory risk.
Ownership
Squads own their product metrics and their on-call.
Innovation
A tenth of every sprint is reserved for experiments.
Feedback
Written feedback cycles every six weeks.
Diversity & Inclusion
Structured interviews and published salary bands.
First 30 days
Ship a production change to the disputes dashboard and join three customer calls.
90 days
Report experiment results and make a keep / kill recommendation.
Outcomes
Own a product area and its metrics end to end.