AI Automation Engineer

Fortinet•Sunnyvale, CA

6d•Onsite

About The Position

We're looking for a junior engineer to be the primary engineer building and maintaining our technical interview pipeline - from the question bank to AI-assisted scoring to full workflow automation - working alongside the team to grow this role into a broader automation function over time. This isn't a "call an API and ship it" job: you'll be designing the systems that add/calibrate interview questions and evaluate candidate answers, which means understanding how to prompt and structure AI models for consistency, how to measure and validate output quality against ground truth, and how to catch and correct the ways these systems get things wrong. This is a role we expect to grow - the shape of that growth is still being defined.

Requirements

Bachelor's degree in Computer Science or related field, or equivalent practical experience
Solid programming fundamentals (Python, JavaScript/TypeScript, or similar)
Comfort working with APIs, including LLM/AI APIs (OpenAI, Anthropic, etc.)
Demonstrated experience designing and evaluating prompts/pipelines for LLM-based tools (not just using AI products, but building with them)
Proficient with Git and a hosted platform (GitHub or GitLab) - branching, PRs/MRs, code review, CI basics
Basic understanding of databases (SQL or similar)
Strong attention to detail and technical writing ability
Ability to work independently on processes that are still being defined, and adapt as scope evolves
Must be authorized to work in the U.S. without sponsorship.

Nice To Haves

Familiarity with test/assessment platforms or ATS integrations
Exposure to scripting automation (e.g., workflow tools, cron jobs, CI-style pipelines)
Interest or coursework in fairness/bias in ML systems
Prior experience conducting or coordinating technical interviews (not required, but a plus)

Responsibilities

Add, revise, and retire technical interview questions across roles/levels/skill areas
Maintain question metadata (difficulty, topic tags, expected answer criteria, role relevance)
Create variant/randomized versions of questions to reduce answer-sharing risk
Build validation checks (automated and manual) to catch incorrect, ambiguous, or miscalibrated questions before they enter the repository
Build automation to update problem sets and recalculate average difficulty ratings as new results come in
Review and update questions periodically to prevent leakage/staleness and keep content aligned with actual job requirements
Architect scoring workflows that evaluate and compare candidate answers against reference answers/rubrics, grounded in a clear methodology for what "correct" and "well-scored" mean
Calibrate and benchmark scoring prompts/models against human-graded samples, measuring accuracy and identifying systematic errors or bias
Understand and account for model failure modes (inconsistency, hallucination, prompt sensitivity) and design safeguards around them
Flag edge cases or low-confidence scores for human review rather than fully automating high-stakes decisions
Automate the end-to-end workflow: question selection -> test delivery -> answer collection -> scoring -> reporting
Help transition our source code and repositories to an internal Git platform
Track test results to help establish scoring benchmarks over time
Track scoring consistency and accuracy over time; report on pipeline health
Watch for and mitigate AI evaluation bias across candidate demographics or answer styles
Document the process, prompts, and scoring logic for auditability
Apply the same automation and AI-integration skills developed here to other internal workflows as the role expands
Partner with engineering teams to identify manual, repetitive processes that are good candidates for automation
Take on additional internal tooling and infrastructure projects as capacity and scope grow