Job description

About Radar Radar radar.com is the global leader in geolocation, with geofencing SDKs, maps APIs, and AI-enabled solutions for marketing, fraud, and operations teams. Why is Radar the best place to work? - We're trusted by some of the world's best companies, from high-growth startups to the Fortune 500. - We have incredible scale: We're processing over 1 billion API calls per day from hundreds of millions of devices. - We're well-resourced, and we've raised $85.5M from world-class investors, including Accel and Insight Partners. - We have a high-performance culture, with ambitious and entrepreneurial teammates in every role. - We recently moved into an amazing new office in Flatiron, Manhattan, NYC. - We were recently named a top 10 best place to work in NYC by Crain's. Despite our growth and scale, we're still just getting started. That's where you come in. About the role: We're looking for a Senior/Staff Software Engineer, Search to improve the data and search systems behind Radar's geocoding platform. Our daily traffic is over 1 Billion API calls / day (15,000 requests / second), so you will be operating at scale. You'll improve the quality and coverage of our location data, build and automate the pipelines that keep it growing, and improve how we understand queries, retrieve candidates, and rank results across geocoding and autocomplete. You'll use ML and LLMs where they can materially improve these systems, from data enrichment and matching to query understanding and search relevance. This is a software engineering role first, you don't need a research background, but you should be excited to work across data, search, and backend systems and reach for ML when it's the right tool. This role will be based out of our NYC HQ. How we work: Most of our engineering team are former technical co-founders or former Radar interns from schools like Waterloo and CMU. Most engineers at Radar fit one of two molds, technically: either Staff level expertise in one stack, or "Multi-Stack" at any level. We say "Multi-Stack" because "Full-Stack" has the connotation of "Frontend and Backend", but Radar Engineers might also work on Mobile or Data engineering. Not that you need to be an expert in all of those, but a desire to learn, jump around to different stacks, and get things done is the important part. We care a lot about shipping fast and talking to customers. We're committed to our product vision of full-stack location infrastructure, but we also know that customer feedback is a treasure map to gold. Even though Slack is the brain of our company, working together in-person in our NYC HQ is the fastest way for us to get things done. We meet on Mondays to plan out work for the week in small groups and use Linear for planning. To us, a week is a long time, and we expect to ship big things every week. The stack: Our core backend service is HorizonDB, a geospatial database written in Rust that powers geocoding: forward and reverse geocoding, autocomplete, and address validation. It's built on custom search indexes and stores like Tantivy, RocksDB, and FSTs, with ML models for query understanding and ranking. Our geospatial data pipelines are written in Scala and Python on Airflow, merging commercial and open data sources into the indexes HorizonDB serves. The API layer is a Node.js TypeScript app, and we use MongoDB, S3/Athena, and Redis. Everything is deployed to AWS via Kubernetes on EKS using Terraform. Most engineers are in the on-call rotation. We sponsor OpenStreetMaps, MapLibre, and OpenAddresses. How we use AI: - Engineers choose what AI tools they use, Claude and Codex being the most popular. - We're actively building Claude skills - for example we've taught it how to debug HorizonDB, our geospatial database. - All code changes are reviewed by an Engineer knowledgeable in that area. Claude and Codex also review all PRs. - There is a range of how much engineers use AI. Most use it daily if not weekly. - We are excited about what AI can do, but we also recognize the risks and don't compromise our coding standards. The hiring process: After a call with our Technical Recruiter, you'll do several technical Zoom calls with members of our engineering team: code screen, coding round, and system design round. If those go well we'll invite you to our NYC HQ for a final round interview. You'll meet one of our co-founders, someone from outside engineering, and meet more people from Radar. We'll go into more depth about how we work to see if there is a match. What you’ll do: - Improve geocoding and autocomplete quality through better query understanding, retrieval, ranking, and relevance - Use ML and LLMs to enrich, label, match, and expand our location data at scale - Automate the data pipelines that keep our coverage and detail growing, reducing manual work and making routine corrections easier to handle automatically - Onboard new and messy data sources, improving how we normalize, match, and integrate US and international location data - Work across the boundary between data and search, improving how location data is indexed, retrieved, ranked, and surfaced to users - Range across the stack, from data pipelines in Scala and Python to search and serving in HorizonDB (Rust) - Have your work power over 1 Billion API calls / day - Talk to Radar customers and prospects, hear their feedback, incorporate it into your work, and make them successful You should: - Think of yourself as a generalist engineer first, not tied to any particular stack - Are interested in talking to customers or prospects and making them successful - Are deeply curious about how things work, and have the tenacity to sit with hard problems and power through them Bonus points if you: - Are a former technical co-founder - Have worked on a high throughput, high growth SaaS application - Have worked on geospatial infrastructure - Have applied ML or LLMs to a real product or data problem - Have worked on search, ranking, relevance, query understanding, or information retrieval systems You’ll work with: - Tim Julien linkedin.com, CTO - Rishant Dwivedi linkedin.com, Engineering Manager - Our customers and prospects - Our Customer Success, Sales Engineering, and Sales teams What we offer: - Competitive salary - Meaningful stock options in a fast-growing company - 401(k) plan with 4% match - New HQ in Flatiron, NYC - Top-notch equipment - Catered lunches - Unlimited PTO - Health, dental, and vision insurance with 100% coverage for employees - 12 weeks of paid parental leave - Commuter and fitness benefits We’ll share full details of our benefits package at the offer stage. Benefits may vary by location. Radar is an equal opportunity employer and does not discriminate on the basis of race, color, ancestry, religion, sex, national origin, sexual orientation, age, citizenship, marital status, disability, gender, gender identity or expression, veteran status, or any other characteristic protected by law.

Requirements

Must-haves

To us, a week is a long time, and we expect to ship big things every week.
Our core backend service is HorizonDB, a geospatial database written in Rust that powers geocoding: forward and reverse geocoding, autocomplete, and address validation.
It's built on custom search indexes and stores like Tantivy, RocksDB, and FSTs, with ML models for query understanding and ranking.
Our geospatial data pipelines are written in Scala and Python on Airflow, merging commercial and open data sources into the indexes HorizonDB serves.
The API layer is a Node.js TypeScript app, and we use MongoDB, S3/Athena, and Redis.
Everything is deployed to AWS via Kubernetes on EKS using Terraform.
Most engineers are in the on-call rotation.
We sponsor OpenStreetMaps, MapLibre, and OpenAddresses.
Engineers choose what AI tools they use, Claude and Codex being the most popular.
We're actively building Claude skills - for example we've taught it how to debug HorizonDB, our geospatial database.
All code changes are reviewed by an Engineer knowledgeable in that area. Claude and Codex also review all PRs.
There is a range of how much engineers use AI. Most use it daily if not weekly.
We are excited about what AI can do, but we also recognize the risks and don't compromise our coding standards.
Improve geocoding and autocomplete quality through better query understanding, retrieval, ranking, and relevance
Use ML and LLMs to enrich, label, match, and expand our location data at scale
Automate the data pipelines that keep our coverage and detail growing, reducing manual work and making routine corrections easier to handle automatically
Onboard new and messy data sources, improving how we normalize, match, and integrate US and international location data
Work across the boundary between data and search, improving how location data is indexed, retrieved, ranked, and surfaced to users
Range across the stack, from data pipelines in Scala and Python to search and serving in HorizonDB (Rust)
Have your work power over 1 Billion API calls / day
Talk to Radar customers and prospects, hear their feedback, incorporate it into your work, and make them successful
Think of yourself as a generalist engineer first, not tied to any particular stack
Are interested in talking to customers or prospects and making them successful
Are deeply curious about how things work, and have the tenacity to sit with hard problems and power through them

Nice-to-haves

None flagged.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aiapimlrustpythonnode.jstypescriptawskubernetesgo

Platform: ashby
Effort: quick

Benefits

Benefits detected in the listing.

Dental

Vision

401(K)

Pto

Parental Leave