Your cart is empty.
Keep shopping
The New SDLC: Agentic Engineering for Software Teams
New
34 students
The New SDLC: Agentic Engineering for Software Teams
Design, verify, and govern software built by autonomous coding agents — from harness to review to organizational rollout
Created bySkliar Serhii
Last updated 8/2026
English
English [Auto],
What you'll learn
- Classify any coding agent by control surface and autonomy tier, and decide what work it's safe to delegate versus what still requires a human in the loop
- Run intent-centric code review on AI-generated pull requests — auditing specification conformance and verification evidence instead of reading line by line
- Build a defensible total cost of ownership model for agentic engineering, including the hidden costs most organizations never measure
- Identify and mitigate the new attack surface agentic systems create: prompt injection, tool poisoning, and supply chain exposure through third-party agent tooli
Explore related topics
Course content
10 sections • 36 lectures • 5h 25m total length
Requirements
- Professional software engineering experience (this is not a beginner "learn to code" course, and it assumes you can already read and reason about production code)
- Familiarity with using at least one AI coding assistant or agent (Claude Code, Cursor, Copilot, or similar) — you don't need expertise, just prior exposure
- Engineering leaders and managers: no hands-on coding required for the leadership-focused sections of each lesson, but a working understanding of your team's SDLC will help
Description
This course contains the use of artificial intelligence. Autonomous coding agents can now generate thousands of lines of working code in minutes. That's not the hard part anymore. The hard part is knowing whether any of it is correct, who's accountable when it isn't, and how your organization changes when implementation
stops being the bottleneck.
The New SDLC is a complete operating system for agentic software engineering — built for senior engineers, tech leads, and engineering leaders who need to move past "prompt and pray" and build the verification, review, and governance systems that make autonomous delivery actually trustworthy at scale.
Each lesson includes a video explainer, an audio podcast version, full written notes, pptx presentation, an infographic, and a mind map. Module quizzes and a downloadable knowledge base are included.
This isn't a prompting course. It's a systems course. You'll learn:
- Why "the fluent lie" — AI-generated code that looks perfect and is subtly wrong — breaks traditional code review, and the intent-centric review protocol that replaces it
- How to build the harness: the sandboxes, gates, evals, and observability that let you safely delegate real work to autonomous agents
- The economics of agentic delivery — how to calculate real total cost of ownership, optimize token spend, and build a leadership dashboard that doesn't lie to you
- How to run parallel agent workstreams without producing an integration disaster
- The new attack surface agents create — prompt injection, tool poisoning, supply chain risk — and how to build a security posture that survives it
- How to translate written AI policy into deterministic, machine-enforced guardrails
- What actually happens to team structure, career ladders, and the junior-to-senior pipeline when implementation capacity stops being the constraint
- How to roll this out in your own organization without a mandate — using a staged, evidence-led adoption plan instead of a directive that produces compliance theater
Every module ends with a real, reusable artifact — a review protocol, a harness spec, a TCO model, an adoption plan — that accumulates into a complete agentic engineering operating system for your own team by the end of the course.
Built around three recurring case companies (a B2B SaaS company, a fintech payment processor, and a legacy enterprise vendor) so every technique is grounded in a real delivery context, not a toy example.
Central thesis: autonomy isn't a feature of the model. It's a function of verification and harness quality. This course teaches you to build both.
Who this course is for:
- Senior engineers and tech leads who are using AI coding agents daily and want to move from ad-hoc prompting to a real operating discipline — code review, quality gates,
- Engineering managers, directors, and CTOs who need to build the business case
- Platform and DevOps engineers who will own the harness, sandboxes, and gates