SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 185,000 - 245,000 / annual
Citizen is the #1 safety app in the U.S., operating a real-time system that ingests signal from 911 radio, CAD feeds, user video, and partner feeds; detects and verifies incidents; and distributes alerts to millions of devices in seconds. The company is being rebuilt as an AI-native platform, with AI now embedded in the detection and verification pipeline itself.
You will own a core piece of Citizen's backend infrastructure and ship it with a small team of engineers and AI agents. The company operates on a driver-and-builder model: every initiative has one accountable driver who owns scope and tradeoffs, and builders who ship alongside them. Drivers are chosen by who wants the responsibility, not seniority, and you will drive your own initiatives early.
Potential ownership areas include: the real-time pipeline (ingest, detection, verification, geolocation, distribution); live video infrastructure (ingest from phone, restreaming, delivery via Mux); AI in the pipeline (evaluation, provenance, guardrails, staged human-in-the-loop to autonomous operation); internal operator tooling (ProtectOS and Regulator, used by Mission Control); and enterprise/API/partner services.
You will work daily with iOS/Android on contracts, ML on models, Data on events, and Mission Control as your primary internal user. A meaningful share of backend work is tooling for the operations team—they are your most demanding users.
In your first 90 days: read the code and use the product daily in New York (days 1–30); ship to production in week one and join on-call as a shadow; drive your first initiative end-to-end and take your first solo on-call week (days 31–60); ship the first AI-in-pipeline capability or video improvement to real users (days 61–90).
Success is measured on: time from signal to verified alert (improving); uptime and latency during major incidents; live video start time, reliability, and cost per viewer-hour; false positives and missed incidents (both falling); infrastructure cost per incident (falling); enterprise/API reliability; on-call resolution time; and cycle time from idea to production (falling as agent loops carry more work).
The role requires you to be in the New York office every day. On-call is rotational (roughly one week every two months) with 24/7 availability for urgent emergencies and resolution expected in hours.
REQUIREMENTS:
- Built and run backend systems at consumer scale (real-time or event-driven); can point to services you personally owned and what happened when they broke
- Fluent in Go; comfortable in Python for ML and data pipelines; at home on Kubernetes and modern cloud stack (message queues, relational and analytical stores, observability)
- Carried a pager for a system people depended on; have a story about a critical incident
- Experience with live video or streaming infrastructure, or demonstrated ability to learn such systems quickly
- Put ML models or LLMs into production paths; understand the difference between demo and production system (evaluation, fallbacks, audit)
- Already build with AI: have replaced parts of your own workflow with agents, have opinions about where they fail, paying attention to what is next (not optional)
- Measure and instrument before arguing; know the difference between an event that fired and an event that landed in the warehouse
- High agency: handed a direction, come back with a better one; want to drive, not only build
- Care about the mission: understand safety is real and build systems like it matters
