Coming soon

Learn agentic automation by designing agents, not reading about them.

Real-world case studies where you make the architecture calls, then get rubric-scored feedback in seconds. From Parth Doshi and the Learning by Doing YouTube channel.

  • Agentic automation
  • Interview prep
  • UiPath · Claude · MCP
  • Free while in beta

What's coming

A lab, not another question bank

01

"Design this agent" case studies

Realistic enterprise scenarios like invoice exceptions, claims intake and IT ticket triage. You make the calls: API, RPA or agent? Where does a human sign off? What happens when it fails?

02

Rubric-scored feedback in seconds

Submit your design and get scored against a published rubric, criterion by criterion, then compare with an example answer from an experienced architect.

03

Every video has a lab

Each Learning by Doing video on YouTube links to a hands-on exercise, so you practise what you just watched instead of forgetting it.

Visit the YouTube channel ↗

Later Spoken mock interviews, so you can practise explaining your design out loud.

Concept-first

Tools change. Judgment doesn't.

Every lab is built around concepts that outlast any product release. UiPath is the worked example today; LangGraph, the Claude Agent SDK and MCP come next, so you learn the idea once and apply it anywhere.

  • Choosing API vs RPA vs AgentWhen an agent is the right tool, and when it is an expensive mistake.
  • OrchestrationCoordinating agents, robots, APIs and people across a long-running process.
  • Human-in-the-loopWhere humans approve, correct, or take over, and how that feeds back.
  • GuardrailsConstraining what an agent can see, say and do.
  • EvaluationsProving an agent works before and after it ships.
  • Tool designDesigning the actions an agent can take so it uses them correctly.
  • Context & memoryWhat the agent knows, where it comes from, and what it remembers.
  • Document understandingExtracting reliable data from messy, real-world documents.
  • Governance & observabilityAuditability, cost control, monitoring and compliance at scale.
  • Failure & recoveryRetries, fallbacks, escalation and graceful degradation.

Who's building this

Parth Doshi

Agentic automation architect, speaker & mentor

  • Speaker at UiPath Community Mumbai: Agentic Days & virtual sessions
  • Interviews and mentors engineers moving into AI roles
  • Runs the "Learning by Doing" YouTube channel
  • UiPath Certified: Agentic Associate & Professional
  • UiPath Certified Solution Architect
  • UiPath Certified Advanced Professional Developer
  • Claude Certified Associate: Foundations

More about the project →