SULICOM

SULICOM · Agentic Engineering

Vasyl Borovyk

Workshop - Loop Engineering

Certificate verified
Certificate WS-8RL2YU4E79 — Vasyl Borovyk

Certificate ID

WS-8RL2YU4E79

Issued

September 1, 2026

Workshop date

July 11, 2026

Duration

6 h 1 min

Share your achievement

Add your certificate to your LinkedIn profile or share it with your network.

Add certificate to LinkedIn
Download PNGDownload PDF

About the program

A hands-on live workshop on making AI agent code trustworthy. The problem is not the model but the missing verification between agent steps: without a pass/fail signal, the developer remains the only reviewer. Loop Engineering is a system of checks the agent runs on its own. Format: one day, two parts. Part one is theory: where feedback loops come from, context as a resource, a deterministic loop around the model, the six verification phases, subagents, TDD with agents, Definition of Done and the self-improvement loop. Part two is live practice on the url-shortener training project: the housekeeper loop, the Ralph loop with safety brakes, a feature built through TDD, eval grading of skills via agent-as-judge and deployment monitoring. Total volume: about 6 hours including participants' Q&A. Participants received access to the recording, the project repository with ready-made verification scripts and environment setup instructions.

Instruction language: Ukrainian

Learn more about the workshop →

Skills

Curriculum

  1. 01Part 1 - Theory

    • Why agents are hard to trust and where feedback loops come from
    • Evolution of approaches: autocomplete, prompt, harness, loop
    • Context as a resource: lost in the middle and context rot
    • A deterministic loop around the model: Ralph loop, Goal, model-as-judge
    • The six phases of the verification loop
    • Subagents and agent teams inside the loop
    • TDD with three agents
    • Definition of Done and the self-improvement loop
  2. 02Part 2 - Practice on the url-shortener project

    • Environment setup: doctor, verify, dev
    • The housekeeper loop: zero-token checks
    • The Ralph loop with safety brakes
    • Implementing a feature through TDD
    • Eval grading of skill behaviour: agent-as-judge
    • Deployment monitoring: Playwright, logs, metrics
    • Q&A: participants' questions

Need help with your certificate?

If you noticed an error in your details or have a question about the certificate, contact our support.

Contact support →