Code Four is an AI‑first public‑safety startup that transforms hours of vision footage (think bodycams, security cameras, LPRs) into context in seconds, giving municipal, county, state and federal agencies back thousands of administrative hours.

Skills: Next.js, TypeScript, GCP, Distributed Systems, Realtime Video/Audio, GraphQL/REST

We’re looking for a fullstack engineer to own massive slices of our product surface: from low‑latency video ingest to the interfaces officers use to review evidence and approve auto‑generated reports. You’ll work shoulder‑to‑shoulder with the founders, ship fast, and help define the engineering culture and architecture that powers mission‑critical workflows for law enforcement.

💻 What You’ll Own

• Event‑driven backend: Architect our Next.js / TypeScript / GCP stack. Design schemas, queues, and webhook ingestion from BWC vendors, RMS/CAD, and DEMS to drive real‑time, personalized officer workflows.
• ML integration surface: Productionize CV / VLM / ASR endpoints from our ML team and expose them through resilient APIs with strict latency/error budgets.
• Evidence UX: Build intuitive dashboards that let officers review, redact, and approve reports with minimal clicks—while preserving full auditability.
• Security & compliance: Enforce end‑to‑end encryption, audit logging, role‑based access, and CJIS controls across services.
• System design & technical strategy: Decide how we scale to dozens of agencies, each with unique policies and templates, without sacrificing speed.

We've built multi‑modal machine‑learning pipelines — CV, VLMs, and automatic speech recognition models built in PyTorch and TensorFlow — that detect objects, track scenes, caption footage and transcribe audio with sub‑second latency.

We’re building a multimodal system that can contextualize any law‑enforcement data source in real time.