Own the implementation and inspect each change
You define the outcome, write or supervise the patch, run the checks and decide whether the result is ready.
Book a Call Self-paced training for software engineers
A practical self-paced course for software engineers moving from agent-assisted coding to governed agentic delivery: authority, context and memory, loop engineering, verification, routing, controlled roles and evidence-gated operation.
What changes
Follow one registration change from supervised agent work to a bounded delivery loop. You will design its authority, context and memory, verification, model routes, role boundaries and evidence-gated path to production.
You define the outcome, write or supervise the patch, run the checks and decide whether the result is ready.
The loop selects bounded work, recalls trusted context, runs in isolation and advances only when the evidence satisfies rules you designed.
Course path
The core videos run for about 54 minutes. Each lesson connects to the same runnable example, a practical lab and reusable worksheets so the operating model can be applied to your own codebase.
See the shift from using an assistant to designing a governed delivery system, and choose the production-adjacent workflow you will improve through the course.
4 minSeparate autonomy from authority, classify risk and define a deterministic shell around the agentic core.
5 minBuild finite context packs, connect Rembr to the runtime lifecycle, and choose between direct context, RAG and bounded recursive processing.
10 minEngineer the operating loop outside the model: triggers, work selection, isolation, fresh runs, verification, persistence and terminal outcomes.
8 minUse deterministic gates, fresh-context review and merge-readiness evidence to decide whether agentic output is engineering work.
7 minDefine stable model lanes, route by task and data class, and measure whether the routing policy is actually economical.
5 minSeparate orchestration, implementation, review and merge responsibility while keeping every handoff explicit.
5 minDesign gates that issue the merge verdict, then review the running product and route defects to repair or source correction.
7 minTurn run evidence into a practical 30-day adoption plan and improve the workflow without mistaking activity for value.
4 minOptional implementation detail for running and verifying the companion repository.
3 min · optional appendixWhat you will be able to do
Separate autonomy from authority and classify delegated work by risk and blast radius.
Build finite context packs, RAG retrieval and durable Rembr memory around the model call.
Design bounded delivery loops, including Ralph-style fresh runs and explicit stop conditions.
Choose verifiers and make deterministic evidence issue the engineering verdict.
Route work across local, hosted and specialist model lanes.
Split delivery across controlled roles without losing merge ownership.
Design evidence-gated merge, product review and remediation routes.
Operate and improve an agentic workflow using evidence from real runs.
Start now
Pay once, register one named learner, and work through the course at your own pace. Your progress is saved so you can return whenever you need it.
Invited by your organisation?
If Radical Geek has approved your email address or organisation domain, register here. You still get your own named account and your own progress record.
Need this for a team?
Use the form to discuss team access, a live working session, or applying the operating model to one of your delivery workflows.