About the company
Ex-Airbnb team building an MVP at the intersection of consumer digital mental health, wearables, and community – currently in stealth.
Role overview
We're looking for a backend engineer with deep data science experience to own the data layer. Work will initially focus on fundamentals (infra, data integrity, reliability) and grow quickly toward ML modeling on top of it.
Day to day:
Building and refining the data pipeline that ingests data from various sources
Deriving clean, meaningful metrics from raw data — aggregating, categorizing, and computing behavioral patterns from noisy inputs
Implementing and extending composite scoring algorithms that combine multiple signals into intuitive user-facing scores
Structuring data so an LLM-powered feature can quickly read a user's full picture and give contextual, personalized output
Building ML models to personalize the user experience — predicting which content is most engaging for a given user, detecting anomalies in patterns, identifying correlations users wouldn't spot on their own
Exploring lightweight recommendation and ranking systems across a large and growing content library
Improving API performance, background job processing, and data integrity across the pipeline
Tech Stack
Backend : Type Script, Bun, Hono
Database : Postgre SQL, Prisma ORM
Background jobs : pg-boss
Auth : AWS Cognito
AI : Claude
Infrastructure : AWS EC2, Code Deploy, S3
Monitoring : Sentry, Post Hog
You
Strong backend experience with Type Script or similar (Python, Go, Java)
Comfortable working with data pipelines: ingesting, transforming, aggregating, and scoring structured data
Experience with relational databases (Postgre SQL preferred) and ORMs
Hands-on experience with ML — even if it's lightweight: recommendation systems, ranking models, anomaly detection, time-series analysis, or personalization.
Familiar with personal user data, time-series data, or data-heavy consumer products (a plus, not required)
Self-directed: you'll get clear specs and context, but you own execution end-to-end
Looking for something long-term, not a quick contract. We want someone who'll grow with the product.
Commitment & Compensation
Fully remote, contract position. Full time (50+ hours/week), able to overlap at least 3-4 hours with US Pacific time for syncs
