About Obsidian Security

Obsidian helps companies figure out what’s happening in their SaaS applications - who’s accessing what, where data is going, and what looks suspicious.

Role Description

You’ll work across the full stack on the Threat product team: browser extension code that hooks into web applications, backend services processing millions of events, and data pipelines feeding our detection engine. One day you’re in TypeScript debugging why a content script isn’t capturing form submissions on some vendor’s weird SPA. The next you’re in Python fixing a Kafka consumer that’s falling behind. Sometimes you’re in Rust optimizing a hot path in the telemetry collector.

Right now, a big focus is shadow AI - enterprises want to know when employees are pasting sensitive data into ChatGPT, Claude, or whatever new LLM chatbot showed up this week. You’ll build the systems that catch this.

What you’ll do

• Own features from “we need to detect X” through production, across extension, backend, and pipeline code
• Write content scripts that interact with web pages and background scripts that coordinate everything
• Build backend services: event ingestion, enrichment, alerting
• Work with security researchers to turn threat intel into working detections
• Help the team get better through code review and design discussions

What we’re looking for

• 8+ years building software, ideally touching a few different areas (not just one stack forever)
• You actually understand how browsers work. DOM APIs, the event loop, how SPAs route, why CORS exists. Not just “I used React.”
• You’ve built browser extensions before. You know the content script/background script split, message passing, manifest v3 limitations.
• Strong TypeScript. Comfortable in Python. Willing to write Rust (or already do).
• You’ve dealt with data at scale: event streaming, pipelines, high-throughput ingestion.
• Backend basics: APIs, Postgres, Elasticsearch, Kafka or similar.
• Can work without someone telling you what to do next.

Bonus points

• You’ve thought about AI security - prompt injection, data exfiltration, that kind of thing
• Background in detection engineering, SIEM, or security ops
• You’ve built or contributed to security tools