Member of Technical Staff, AI Engineer

Parameter ai; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Parameter ai in San Francisco is seeking a Member of Technical Staff, AI Engineer to build LangGraph/Temporal-based agent systems for autonomous offensive security testing. You will design and ship complex agent workflows, implement the tool layer, and own the evaluation loop to drive better results while reducing false positives at scale.

The role requires strong TypeScript and experience with async, distributed systems, long-running workflows, and context-aware tool design.

Qualifications

• Minimum 3+ years of software engineering experience.
• At least 1 year building production LLM systems.
• Strong TypeScript proficiency and async skills.
• Experience with distributed systems and long-running workflows.

Responsibilities

• Design and ship agent workflows in LangGraph and Temporal to plan, execute, and verify multi-step exploitation chains.
• Build the agent tool layer, including HTTP clients, auth handling, browser control, cloud enumeration, and payload generation.
• Own the evaluation loop to define what good looks like and improve outcomes over time.
• Work to reduce false positives and address top customer priorities.
• Manage inference cost and latency at scale with prompt caching and model selection across Claude and AWS Bedrock.

Skills

TypeScript

Async development

Distributed systems

Long-running workflows

Context engineering

Tool design

Eval harness

Security awareness

Tools

LangGraph

Temporal

GCP

AWS Bedrock

Anthropic API

GitHub Actions

Linear

Graphite

Job description

Parameter ai is hiring a Member of Technical Staff, AI Engineer in San Francisco to build LangGraph/Temporal-based agent systems for autonomous offensive security testing.

Responsibilities

• Design and ship agent workflows in LangGraph and Temporal to plan, execute, and verify multi-step exploitation chains
• Build the agent tool layer, including HTTP clients, authentication handling, browser control, cloud enumeration, and payload generation
• Own the evaluation loop by defining what “good” means for agent runs using Braintrust and Langfuse, with the goal of improving outcomes over time
• Work to reduce false positives, identified as the hardest problem in this space and a top customer priority
• Manage inference cost and latency at scale, including prompt caching, context strategy, and model selection across Claude and AWS Bedrock

Requirements

• 3+ years of software engineering experience
• At least one year building production LLM systems (not prototypes or demos)
• Strong TypeScript
• Comfort with async development, distributed systems, and long-running workflow orchestration
• Intuition for context engineering, tool design, and why agents fail in real environments
• Data-driven evaluation: built or maintained an eval harness, including understanding why offline metrics can diverge from production behavior
• Bonus: security background, CTF experience, bug bounty history, or a track record of breaking things you were not supposed to break

Technologies

• TypeScript, LangGraph, Temporal
• GCP
• AWS Bedrock, Anthropic API
• Langfuse, Braintrust
• GitHub Actions, Linear, Graphite

Interview Process

• Intro call with a founder (30 minutes)
• Technical deep dive on your past work (60 minutes)
• Paid work trial or take-home, scoped to roughly one day
• Onsite in San Francisco with the team
• References and offer

Location and Work Authorization

• Onsite in San Francisco, CA
• US work authorization required; visa sponsorship is not available

Compensation: USD 180,000 - 240,000 per year