Presented by
NYCOO
Prepared for Masttro

Practical AI

From mindset and prompting, to agents and connected tools, to building custom applications and adopting AI across your organization.

Format
Four sessions, one hour each
Delivery
Live, with demos and hands on work
Prepared for
Masttro
Program Overview

AI is as fundamental to modern work as email, documents, spreadsheets, and the internet. This program makes that real for your team.

The goal across these four sessions is to take your team from knowing about AI to using it as a default tool, then to building with it, and finally to adopting it across your organization.

Each session is one hour, delivered live, virtually, with demos, and hands on work. Every session builds on the one before it, and each ends with a homework task so the time between sessions is productive.

What your team walks away with

  • A working mental model of what AI is, what it is good at, and what to be cautious about.
  • A practical command of prompting that applies across every AI tool.
  • Fluency in chat tools beyond the basics, including Projects, memory, customization, and tool connections.
  • A real understanding of agents, browser use, and how AI takes action on your behalf.
  • Exposure to custom AI applications and what it takes to build them.
  • A concrete adoption plan with owners, triggers, and incentives.

How the sessions are structured

  • One hour per session, four sessions total.
  • Live demos throughout, not just slides.
  • A short hands on moment in each session, so attendees leave with their own result rather than just notes.
  • Each session ends with a specific homework task that sets up the next session.
  • Scheduled roughly one week apart so the homework has time to land.
01/04
Session One

Foundations, Mindset, and Prompting

Shift how the room thinks about AI. Establish that prompting is the foundation under everything. Each person writes their first real prompt for a real task.

Agenda
  • 0 to 5 min
    Welcome and framing
    What the four sessions cover and how they build on each other.
  • 5 to 10 min
    The mindset shift
    AI as fundamental as a doc or browser tab. Today is the worst AI will ever be.
  • 10 to 15 min
    Audience calibration
    Who uses chat tools. Who has built a Project or Custom GPT. Who already saves real time each week.
  • 15 to 20 min
    State of the market
    Lessons from 70+ sessions. Why adoption is slow. Why internal tools matter. Why security versus flexibility is a false choice.
  • 20 to 25 min
    What AI actually is
    A computer doing things only people used to do. The nondeterministic nature of the output.
  • 25 to 30 min
    User beware
    How it actually works, why it hallucinates, and where human judgment still matters.
  • 30 to 50 min
    Prompting best practices, with live demos
    Clarity, roles, context, examples, output format, asking it to explain, telling it how to think, giving it an out, meta prompting, iteration.
  • 50 to 55 min
    Hands on
    We write some prompts together and review the results.
  • 55 to 60 min
    Close and homework
    Use AI on one real task before the next session.
Take home

A packet of prompting tips and guidelines to run at least one real prompt before Session 2.

02/04
Session Two

Basic Chat at the Limit

Show that a well used chat tool, customized and connected properly, already replaces a meaningful share of daily work. Each attendee builds their first Project.

Agenda
  • 0 to 5 min
    Recap and homework share
    What worked and what didn't when prompting.
  • 5 to 20 min
    Core use cases, with live demos
    Content creation and editing. Problem solving and thought partnership. Data analysis. Legal assistant.
  • 35 to 45 min
    Customization
    Privacy and training settings. Custom instructions. Memory.
  • 20 to 35 min
    Claude / ChatGPT deep tour
    Picking the right model. Search. Canvas. Image generation. Voice and video. Working with desktop apps. Tasks on a schedule.
  • 45 to 55 min
    Projects, live build
    Each attendee sets up a real Project for an ongoing task such as a research stream, performance reviews, or a recurring client workstream.
  • 55 to 60 min
    The wider market
    Review of the main tools: Claude, ChatGPT, Perplexity, Copilot, Gemini.
Take home

A working Project on your own account, configured with context and instructions for a real ongoing task.

03/04
Session Three

Beyond Basic Chat. Context, Tools, and Agents

Adding Context and Tools is a fundamental step in getting more value from AI.

Agenda
  • 0 to 5 min
    AI strategy framing
    Top down and bottom up views of your workflows. Where to inject AI. Building in stages. What AI is actually good at.
  • 5 to 10 min
    The three pieces beyond basic chat
    Context plus tools. Customization. Interfaces.
  • 10 to 15 min
    Context and tools explained
    New things to know, such as live data and internal metrics. New things to do, such as updating databases, sending messages, and taking real world actions.
  • 15 to 25 min
    Connections and integrations, live demo
    Connect ChatGPT or Claude Cowork to files, email, calendar, and CRM. Run a real cross system query.
  • 25 to 35 min
    Deep Research
    CRM research, competitor mapping, background checks, sales prep. Attendees run one on a real topic before the session and bring the result.
  • 35 to 45 min
    Browser use, demo
    Full AI browsers such as Atlas, Dia, and Comet. Agent modes and extensions. Real examples including license checks, cleaning broken links, and research workflows.
  • 45 to 50 min
    Computer use and interactive outputs
    AI driving your computer. Interactive data artifacts, dashboards, and AI generated design.
  • 50 to 55 min
    Assistants and agents
    One prompt that does it all. Running multiple agents in parallel while you work on something else.
  • 55 to 60 min
    Close and homework
    Wire up one real connector. Run Deep Research on something that matters.
Take home

At least one connected tool running on your account, and one completed Deep Research report relevant to your work.

04/04
Session Four

Custom Apps, AI Software, and Adoption

Move from using AI to building with AI. Walk through real in house examples. Survey the AI first software category. End with a concrete adoption plan your team can run.

Agenda
  • 0 to 3 min
    Recap
    One connector or agent each person used since last session.
  • 3 to 8 min
    Three ways to build custom
    Custom Chat. Custom apps with visual builders. Custom apps with code. Same three ingredients each time: context, tools, interface.
  • 8 to 18 min
    Custom Chat deep dive
    ChatGPT Projects and Custom GPTs. Claude Projects and Claude Skills. A live walkthrough of a real Custom GPT or Claude Skill built with company knowledge.
  • 18 to 28 min
    Custom apps with code, tour only
    Real examples including a research agent, a collections agent, a news agent, contracts to database, an AI lawyer for contract redlines, and a fully AI built web application.
  • 28 to 43 min
    AI software category, fast survey
    Candidate sourcing and market mapping. Google Workspace with Gemini. Microsoft 365 with Copilot. Claude in Excel. Note takers. Multimedia studios. Natural language app building. Vertical platforms such as Hebbia and Harvey.
  • 43 to 53 min
    Organizational adoption
    The four levers. Train and support. Create the culture. Add incentives. Implement triggers. And assign an owner.
  • 53 to 58 min
    Staying current and action plan
    A short list of podcasts and newsletters worth following. Each attendee leaves with a plan of action.