Senior Analyst of Attribute Analytics (Imagery & Property Data)
Posted Updated
Description
- We’re looking for a Senior Analyst, Attribute Analytics to help us learn what imagery can tell us about a home. Attribute Analytics makes sure Kin has the best property data in the industry at the most cost-effective price, working closely with Underwriting, Pricing & Risk Management (PRM), Product, Engineering, and Finance. The team owns data confidence (the models that decide which property details Kin can trust automatically and which need manual review), vendor procurement and cost, and attribute data quality and value. Imagery is the newest part of that work
- In this role, you’ll run point on what Kin can learn from real estate listing photos and records, aerial imagery, and street-level imagery. You’ll decide which property attributes are worth pulling, build and validate the models that pull them, and prove whether they’re accurate and worth what they cost. You’ll partner closely with a dedicated engineer who brings your work into production. This role is ideal for someone who pairs modeling fundamentals with real business sense, uses AI tools every day to move faster, and knows exactly where those tools go wrong
- Build and validate models that turn imagery into property attributes (exterior finish, roof shape, pool, condition, and more), combining vision model APIs with traditional predictive modeling and using AI coding tools to move fast
- Test those attributes against manual review outcomes and other vendors, and decide how much Kin should trust them and how they feed our data confidence models
- QA your own work and others’ rigorously: confirm AI-assisted code and analysis does what it claims, catch leakage and silent failures, and keep results reproducible
- Find, test, and recommend new imagery and property data sources, weighing coverage, accuracy, cost, usage rights, and build vs buy
- Measure which attributes actually move loss ratio, so Kin invests in the data that matters
- Use imagery to check attribute quality where no manual review exists, and to spot property changes after a policy is bound
- Shape with Engineering what gets built, in what order, and how it’s monitored once it’s live
- Translate findings into short, clear written recommendations for Underwriting, PRM, Product, and leadership, including how confident you are and what could be wrong
- Champion a culture of continuous learning by sharing analytical and AI best practices across the team
- Success in this role:
- In your first 6–12 months at Kin, success is less about checking boxes and more about the impact you create. You’ll use your skills and judgment to take ownership of meaningful work, improve how we operate, and help move Kin’s mission forward. Along the way, you’ll deliver outcomes that make a real difference for both Kinfolk and the homeowners we serve
- By the end of your first year, you should feel confident in your role, trusted as an owner, and proud of the progress you’ve helped make
- Kin has a clear, evidence-backed view of which imagery-derived attributes it can trust automatically and which still need manual review
- Data sourcing and vendor decisions are being made with confidence in what each source costs, covers, and delivers
- Your validated models are running in production with monitoring in place
- Underwriting, PRM, Product, and leadership are using your recommendations to decide where Kin invests in property data
Benefits
- Medical
- Dental
- Vision
- Life insurance
- Disability insurance
- Commuter benefit plan
- 401k
- Flex spending accounts
- Support for continuing education and professional development
- Training, development, and networking opportunities via Sales Assembly
- Flexible PTO
- Onsite gym membership
- Company-provided lunch each month
- Daily snacks
- Sodas, unlimited flavored water, & coffee
- Daily use of AI tools (LLMs, AI coding assistants like Claude Code or Databricks Genie, model APIs), with a clear sense of their pitfalls: hallucinated logic, untested code, overconfident outputs, and results that look right but aren’t
- Prior experience in insurance, insurtech, fintech, or property data
- SQL and Python experience working with large, messy, semi-structured data, such as JSON from third-party vendors
- Experience with computer vision, image data, or multimodal AI models
- Experience independently leading complex, ambiguous projects from scoping through delivery
- 5+ years of experience in data analytics, data science, or a related analytical role
- Experience with geospatial data or aerial imagery
- Hands-on experience building and validating predictive models and measuring them against ground truth (accuracy, coverage, lift)
- Experience translating technical findings into clear recommendations for business partners, and making assumptions explicit when data is incomplete
Skills
Sign in to read and join the discussion.