Structured training
LIVE TRAINING
Your Next Skill:
LLM Test Automation
Learn to test AI-powered applications and automate evaluations. For QA professionals and complete beginners. No coding experience required.
Next cohort: November 10 · Live online lessons · Small cohort
Assert. Compare.
Improve.
Make quality visible with repeatable evaluations.
01ScenarioWhat should the assistant do?
02Promptfoo testsRun checks on each version.
03ReportExplain the failure with evidence.
PROGRAM OVERVIEW
Seven weeks. Real practice.
2 hours each
Small cohort
Career coaching
WHY TEST AI?
AI moves fast. Mistakes cost millions.
Companies are racing to adopt AI, but hallucinations, security breaches, and unreliable responses create serious business risks. That's why AI testing matters more than ever.
ONE TOOL. REAL SKILLS.
Tool overload isn't a curriculum.
Other bootcamps sell you a long list of tools and call it a curriculum. One tool per week? That's barely enough time to scratch the surface. We don't teach tool collecting. We teach LLM test automation. Master Promptfoo through real AI testing scenarios and build a portfolio that proves what you can do.
THE TOOL-OVERLOAD APPROACH
Estimated learning time
600+ hours
Across the full tool stack
THE EVALVED APPROACH
Promptfoo.
One core evaluation framework, supported by VS Code and GitHub.
Estimated learning time
60–80 hours
One focused AI testing workflow
YOUR PORTFOLIO PROJECT
Don't just learn AI testing. Prove you can do it.
Build a complete GitHub portfolio showcasing automated LLM tests, evaluation datasets, regression coverage, documented defects, and a final AI quality audit.
- Automated test suites
- Reproducible defect reports
- Final quality assessment
llm-testing-portfolio
Project Previewtests/Automated test scenariosdatasets/Evaluation datasetsreports/Results and defect evidencedocs/Test strategy and quality auditpromptfooconfig.yamlPromptfoo evaluation configurationREADME.mdPortfolio overview
VIDEO TESTIMONIALS
Real experiences. In their own words.
1–3 of 3
CURRICULUM
Fourteen lessons. Five clear stages.
Learn to test AI for accuracy, reliability, safety, and conversational context. Build Promptfoo checks, interpret results, and complete your final audit.
1Understand the systemWeek 1 · See how AI applications work and what can fail.
- Lesson 1 · Introduction to LLM Test AutomationExplore a practice AI chatbot manually; distinguish variable wording from a quality defect.
- Lesson 2 · LLM Failure Modes, Test Oracles, and RiskStudy chatbot anatomy and failure layers, then turn requirements into risks and expected behavior.
2Build reliable automationWeeks 2–3 · Set up Promptfoo and grow a reusable suite.
- Lesson 3 · Promptfoo Setup and Technical FoundationsSet up on a Mac in a live workshop, trace application layers, and run a guided preflight evaluation.
- Lesson 4 · Building Your First Promptfoo Test SuiteWork in a company-style repository with GitHub and VS Code; add basic assertions and compare two targets.
- Lesson 5 · Reliable AssertionsChoose checks that catch meaningful errors without failing on harmless wording changes.
- Lesson 6 · Datasets, Variables, and Test CoverageTurn scenarios into a reusable data-driven suite and map cases to risks and quality dimensions.
3Judge and compare resultsWeek 4 · Combine human judgment with repeatable evidence.
- Lesson 7 · Human Review and LLM-as-JudgeWrite a bounded rubric, label examples, and investigate disagreement with model-graded checks.
- Lesson 8 · Regression Testing and Result AnalysisCompare stable and buggy versions, rerun uncertain cases, and explain what the evidence supports.
4Test conversations and riskWeeks 5–6 · Investigate state, failure layers, and safety.
- Lesson 9 · Multi-Turn Conversation TestingTest follow-ups, preferences, context carryover, and reset behavior in short conversation flows.
- Lesson 10 · Architecture-Aware DebuggingWorkshop a defect: compare response, data, and run evidence; separate the symptom from a cause hypothesis.
- Lesson 11 · Security and Focused Red TeamingProbe prompt injection, privacy, and unsafe answers; manually review flagged cases.
5Plan, audit, and presentWeeks 6–7 · Turn cumulative work into a defended portfolio.
- Lesson 12 · Evaluation Planning and ReportingSet scope, choose cases, outline the report, and define evidence for a release recommendation.
- Lesson 13 · Shared Project Audit SprintRun the suite, triage failures, review the draft package, and address feedback.
- Lesson 14 · Portfolio Packaging and Career TranslationPresent the audit, defend the recommendation, and explain your contribution clearly.
LIVE LEARNING. REAL SUPPORT.
Learn from practitioners. Not prerecorded playlists.
Lead Instructor
Perice Pope
Cloud and software engineer
Perice brings cloud architecture and software engineering into the classroom, guiding students through hands-on LLM testing.
Instructor & Mentor
Patrick Siebenthal
AI and VR quality assurance
Patrick draws on over eight years in QA, including AI and VR testing for Meta and Intel, to connect testing judgment with practical evaluation.
Founder & Mentor
Yuri Sokolov
QA leadership and technical education
Yuri combines QA leadership experience at Google with curriculum development and teaching, helping students turn quality problems into practical tests.
PRICING & SCHEDULE
Know what you're joining.
14 live lessons from November 10 through December 17, 2026, at 5:30–7:30 PM PST / 8:30–10:30 PM EST. Onboarding is November 9. The first cohort has at most 20 students.
Save $300 on the $1,095 public tuition with an eligible referral. Ask us to confirm eligibility before enrollment.
Ask About Referral PricingLive instruction, guided practice and homework, one portfolio project, individual feedback, and three individual career-coaching sessions.
Enroll at $1,095Participation requires a Mac.
1Who is this program for?
QA professionals who want AI testing skills and complete beginners who want a guided introduction are both welcome.
2Do I need coding experience?
No. We start with the basics and guide you through the required setup. You will need a Mac and familiarity with common computer applications.
3Is the program live or self-paced?
Live. The course is taught through scheduled live lessons, and practice and homework build on what you learn in class.
4What will I have to show after the course?
By the end of the course, you will have a GitHub repository with a portfolio that you can show during interviews. It will include your automated tests, datasets, quality audits, and evidence of defects found.
5Does the program include an internship?
No. The program does not include an internship or job guarantee.
6What is included in career coaching?
Three individual career-coaching sessions cover positioning your skills, presenting your experience, and preparing for interviews. Career coaching does not guarantee employment.
7How much does it cost?
Tuition is $1,095, or $795 with an eligible referral.
8What is the refund policy?
Tuition is non-refundable, except where required by applicable law. Full enrollment terms will be provided before payment.
9Will I have access to lesson recordings?
Our planned commitment is 12 months of replay access from the cohort start date.
10Are Slack support or a certificate included?
Every enrolled student receives an EVALVED certificate of completion after the final lesson. Student support is available at support@evalved.com during the course. The communication platform will be provided during onboarding; Slack has not been confirmed.
A PERSONAL INVITATION
Let's talk about your next step.
Yurii SokolovFounder, Evalved Academy
Book a Free Call
Share your details. We'll contact you to arrange a time.
We'll use your details to respond to this request and arrange a call. Submitting does not subscribe you to unrelated marketing, automated calls, or SMS campaigns. Privacy Policy