Role Overview
Abnormal AI is hiring a Software Engineer to build and scale the platform behind its Identity Threat Protection product, blending backend systems development with ownership of features, production systems, and AI-powered agentic components. You will partner with senior engineers to improve reliability, reduce latency, and accelerate release cycles.
Responsibilities
- Design, build, and iterate on scalable backend and ML systems, APIs, frameworks, and internal tools.
- Own well-scoped features and components, with guidance on complex cross-system work.
- Improve stability, reliability, and operational excellence of critical systems; participate in on-call rotations.
- Write clean, testable, resilient code with attention to edge cases and performance.
- Design and build LLM-powered features and agentic workflows (investigation, triage, remediation assistants).
- Integrate LLM APIs and agent frameworks into backend services, focusing on reliability, cost, latency, and evaluation.
- Contribute to prompt design, tool/function-calling integrations, and guardrails for AI-driven components.
- Collaborate cross-functionally in an async-first environment; contribute to design docs and code reviews.
Requirements
- 3-5 years of industry experience as a Software Engineer, shipping production backend systems.
- Solid backend proficiency in Python, building and maintaining production systems.
- System design fundamentals and ability to reason about scaling and reliability tradeoffs.
- Working knowledge of relational databases and modern data storage technologies.
- Experience with AWS cloud services (S3, RDS) and deployment practices.
- Understanding of service health, monitoring, and incident response practices.
- Comfortable writing technical documentation and contributing to design discussions.
- Exposure to service-to-service communication (gRPC, Kafka) and caching (Redis) is a plus.
- Familiarity with containerization and orchestration (Docker, Kubernetes, Helm) is a plus.
Nice to Have
- Hands-on experience integrating LLM APIs (e.g., OpenAI, Anthropic) into production applications.
- Exposure to agent frameworks or patterns (e.g., LangChain, LangGraph, tool/function calling, ReAct-style agents).
- Familiarity with prompt engineering, evaluation, and observability for AI-driven features.