Microsoft

Sash Mohapatra

AI educator & Builder | 20 Years in Microsoft in Customer engineering and Sales

Cohort 1

Cohort 2

Most people know AI can help with their work. The hard part is deciding what to hand off, what to keep, and how to stay in control. This worksheet gives you a structured framework to answer those questions with your actual work. You'll identify every repeating pattern in your week, score them on impact and stakes, and see exactly where and&nbsp;how to delegate. From tasks you can fully hand off to ones that need your eyes before they ship.

The AI Delegation Decision Matrix

You will learn how to use AI like the top 1% operators.

Founder riftlab.ai | ex-Microsoft

AI educator & Builder | 20 Years in Microsoft in customer engineering and sales

Applied AI for Knowledge Work

Corporate Affairs

Banking Consultant

Technical Program Manager

Strategy & Operations

You're using AI. Your team is using AI. But nobody's built infrastructure with persistent context, repeatable workflows and connected tools. Your capabilities can only compound when you have a system that automates and executes for you.This course builds that system with Claude Code/Cowork. You'll graduate with a working AI system on your machine, built on your context, powered by reusable skills, and connected to the tools where your work lives.In 4 weeks, you will :<ul><li>Move from doing the work to designing the system that does it. The shift from using AI to orchestrating it</li><li>Build AI that remembers your context across every session. Your role, your projects, your standards. Brief it once, it compounds from there.</li><li>Turn your repeatable processes into skills that execute consistent workflows. Weekly reports, client briefs, research summaries. Described once, invoked on demand.</li><li>Connect AI to where your work actually lives. Your documents, your notes, your calendar, your inbox.</li><li>Delegate work to AI agents and review the results. Define the outcome, let the system execute, focus on what needs your judgment.</li></ul>

Build a Personal AI System in 4 weeks. No Code Required.

The Orchestrator Mindset

Communicating Effectively With AI

Building Persistent Context Across Your Work

Deconstructing Your Work Into AI-Ready Workflows

Building Your Personal Intelligence System

This course teaches you how to build a Personal Intelligence system with Claude Code & Cowork. Both require a pro or max Claude subscription

Claude subscription required

AI-first Knowledge Workers: Professionals with deep expertise who want to capture their processes as reusable AI skills that run on demand.

Founders or Consultants: Independents building AI systems that handle repeatable work so they can focus on high-judgment tasks.

Leaders or Managers: Decision-makers building hands-on AI fluency to evaluate, guide, and scale AI adoption across their teams.

AI Educator and Builder | Ex-Microsoft

Learn the product engineering principles every vibe coder eventually needs

You've built something real with AI coding tools. You've shipped it. And you hit the honest problem: you don't know what you don't know.<ul><li>What are the three layers of your app - frontend, backend, APIs, databases - and how do they talk? </li><li>What's a data model, and why does every reliable product start there? </li><li>How do you test changes in dev before users see them in production? </li><li>Why does your agent keep forgetting context? </li><li>Which decisions are safe on instinct, and which cost you a weekend if you guess wrong?</li></ul>These are the engineering questions nobody covers in a vibe coding tutorial. Product managers absorb them through proximity to engineers. Growth marketers, operators, designers, and business-background founders rarely get that proximity - but they need the same answers.This workshop gives them to you. You'll work with AI tools the way someone with engineering intuition does - understanding what you're building and spotting problems early.One day. Four stages: THINK → MAP → BUILD → SHIP. You'll build a working app and a vocabulary that works with Replit, Cursor, Claude Code, or Lovable - the same process I use to ship Pollzy.

Non-technical founders building products on Replit, Lovable, or Cursor who have moved past the prototype stage.

<code>Growth marketers, operators, and BD pros who've shipped internal tools with AI but hit walls maintaining or extending what they built.</code>

Designers, customer success leads, and business-background builders using AI coding tools who want to understand the technical foundations.

Anthropic's new agent for knowledge work - modes, features, real workflows

Founder riftlab.ai | Ex-Microsoft

Claude Cowork is Anthropic's new agent for knowledge work. It runs inside the Claude desktop app, reads files on your computer, and completes multi-step tasks end to end - drafting documents, organizing folders, building spreadsheets, formatting reports.This workshop is a practical tour of what Cowork actually does. I'll walk through the three modes in the Claude desktop app, the agentic loop under the hood, and the features most people don't know exist - Dispatch, scheduled tasks, projects with memory, sub-agents. Then I'll run three real workflows from my consulting work, live, on my own machine.By the end you'll have a clear picture of where Cowork fits in your week, where it doesn't, and how you can use it to save hours in your work week.

Knowledge workers looking to understand how Cowork fits their actual work

Consultants, founders, operators, and managers with weekly drafting, summarizing, and assembly work they'd like to delegate

Cowork early users who've installed it and run a few tasks, and want to get more out of it.

Your CLAUDE.md: The File That Makes Claude Code Yours

Built for coding, powerful for knowledge work orchestration. See how it runs consulting, writing, and marketing work.

Claude Code as a Intelligence system for knowledge work

One markdown file - your identity, voice, standards, and rules of engagement. Loads in Code and Cowork.

CLAUDE.md: the file at the center of the system

See the real files from my Claude Code setup and get prompts to build your own.

Inside the CLAUDE.md files I use for real professional work

The Orchestrator Mindset: Principles for working with AI

Define the Outcome First. Deconstruct, then Delegate. Meta-Plan with AI. Build Systems, Not Outputs.

A four-principle framework for delegating real work to AI

One task - decomposed, delegated to AI, and delivered, using all four principles in real time.

The four principles applied to a real pro task, live

See how the four principles scale from a single prompt to a multi-agent system.

Where agents fit and how to orchestrate them

Setup OpenClaw securely with Claude Code

Founder of Rift Media | AI educator and Practioner | Ex-Microsoft

Live audit of a fresh Hostinger instance : what's listening, how credentials are stored, what's exposed.

What a default OpenClaw deployment actually looks like

SSH, firewall, Tailscale, credential lockdown, brute-force protection. Directed in plain English, executed by AI.

Using Claude Code to secure OpenClaw - in plain english

Wire up a Discord bot with pairing approval so only you get access from anywhere.

Connect Discord to talk to your OpenClaw assistant

Scale Knowledge Work with Claude Code

Founder of Rift Media |  AI educator and Practioner | Ex-Microsoft

The architecture behind an AI system that executes for you, not just advises. No coding required.

The Intelligence Framework: Think, Remember, Act, Use Tools

Watch your team of agents execute your projects, in your voice with your preferences. Full content pipeline, automated.

Research → Write → Draw → Publish in one workflow

Live demo of an agentic workflow that analyzes positioning, features, and weaknesses - work that normally takes half day

Sash Mohapatra will transform you into a systems thinker

Sash Mohapatra

About Sash

Courses

Applied AI for Knowledge Work

Workshops

Product Engineering Foundations for Vibe Coders

Claude Cowork Masterclass : From Curious to Capable

Lightning Lessons

Your CLAUDE.md: The File That Makes Claude Code Yours

The Orchestrator Mindset: Principles for working with AI

Setup OpenClaw securely with Claude Code

Scale Knowledge Work with Claude Code

Resources

The AI Delegation Decision Matrix