Staff AI Security Research Engineer

Alkonos, Inc.; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Competitive salary

Meaningful equity

Visa sponsorship

Job summary

Wakefield is a San Francisco–based applied research lab building AI-native security services. You will own how the agents get better—the benchmark, the eval harness, the validator that proves each finding, and the loop that turns misses into verified improvements.

When a new frontier model arrives, you’ll determine what it can do against real applications, running rigorous tests and turning results into robust improvements. In-person role with visa sponsorship and equity.

Qualifications

• 2+ years of engineering experience or equivalent depth from research.
• Experience at a fast-growing startup is preferred.
• Shipped applied ML systems from scratch as an engineer.
• Ran evals and experiments rigorously against production systems.
• Strong fundamentals: ML evaluation, experimental design, statistics, LLM agents.
• You use coding agents daily and understand every line you ship.
• You build and tinker for fun because you care about how the technology works.

Responsibilities

• Benchmark apps with known vulnerabilities and evals that measure changes to the agents.
• Develop agents that learn across users and tenant boundaries.
• Maintain the validation agent that replays findings in a real browser.
• Turn production traces into testable cases and rerun experiments.
• Run every new model through the harness and evals within days of release.
• Credit assignment over long scans to identify effective bug-finding steps.

Skills

ML evaluation

Experimental design

Statistics

LLM agents

Coding agents

Tools

Harness

Job description

Wakefield is a San Francisco–based applied research lab building AI-native security services. You will own how the agents get better—the benchmark, the eval harness, the validator that proves each finding, and the loop that turns misses into verified improvements.

When a new frontier model arrives, you’ll determine what it can do against real applications, running rigorous tests and turning results into robust improvements. In-person role with visa sponsorship and equity.