AI Fluency: From Prompting to Agent Orchestration

This class is live online (i.e. synchronous).

Is this course for you?

AI Fluency: from Prompting to Agent Orchestration is designed for non-technical professionals. This class is for you if:

  • You're an independent creative, solopreneur, entrepreneur, or part of a start-up team or small business.
  • You've prompted a chatbot such as ChatGPT, Claude, Gemini, or Grok to draft an email, research a prospect, or edit a report, and you want to know what comes next.
  • You're don't just want to know about agentic AI, you're ready to use it and want to how with and from peers and a live instructor.
  • You're thinking about making the leap into entrepreneurship or turning a side project into an automated money maker.

    • This class helps you apply these tools to a specific organization in order to reduce costs and tedious back office drudge work while freeing up more time to do the essential work of the business. That's the orchestration part. You become the conductor, not the one playing every instrument: you set the direction, AI agents play their parts, and the important calls stay with you.

    If you are a non-technical professional or knowledge worker who works in a larger organization or enterprise, you can still gain valuable skills in this program. What you'll learn is all relevant and useful but some topics specific to larger enterprises are not covered and enterprise policies may limit what you can personally directly implement.

    BONUS! Coaching For Your Use Case

    For this first cohort, every student gets a 45-minute session with an instructor to talk through your specific business use case for agentic AI. Bring the work you most want off your plate, and talk through where agents could help. These sessions will be scheduled during the last 4 weeks of the class and the 3 weeks following the class./p>

    What will the course cover?

    • Identifying workflows ready for agentic support
      Which of your current workflows are suited for and ready for agentic support? How to catalog your repetitive tasks, workflows and practices. Identifying whether the data sources you need are available in machine-readable form (your post-it notes won’t help an agent!).
    • The iterative approach to dialing in your agents
      Agents need a “rinse and repeat” NOT “set and forget” approach. Learn to iteratively refine your agents until you get them dialed in to where you want them. And you’ll learn how to monitor performance to ensure they don’t drift.
    • Designing or redesigning processes for agentic support
      Keeping approvals, exception handling, and other key decisions on your plate while moving the repetitive and normal off of it. Learn the design patterns and rules for successful agentic augmentation.
    • Up-to-Date Prompting Guidelines
      The major chatbot vendors are constantly upgrading their large language models - and that changes the guidelines for how to most effectively prompt your chatbot. Learn the newest rules for successful prompting to both better achieve your goals and do so at lower cost.
    • Skills - a foundation for agents
      Skills give your AI tools specific expertise by providing modular, reusable instructions and step-by-step guidance to handle repeatable operational tasks. Skills are one of the key building blocks for agents. You’ll learn multiple ways to define skills based on your own work and processes. We’ll also explore how to find prewritten skills and when and how to safely adapt them for your use.
    • Matching agent platforms to your needs and preferences
      Different agent platforms are better at different things - how do you decide which one (or ones) you want to use? We’ll look at the key architectural differences, and their tradeoffs. We’ll compare differing user interface models, including the agent workbench approach vs. the always-on and team of agents approaches.
    • Creating outputs from your agents
      The vendors have been creating a range of options for creating deliverables, often referred to as artifacts, including dashboards, documents, spreadsheets, slides, video, images, web sites, etc. from their chatbot or an agentic process. You’ll explore how to prompt your chatbot and define skills to create relevant work products.
    • Projects, Workspaces, and Custom Agents
      While vendor terminology constantly shifts—from Claude Projects and ChatGPT custom GPTs to plugins and workspaces—they all solve the same problem: organizing work into dedicated environments. These containers let you pair specialized instructions with tailored document repositories. Instead of rebuilding your context from scratch in every chat, working inside a project environment ensures the AI stays anchored to your specific initiative, files, and operating rules.
    • Context vs. Memory
      These concepts relate to what an LLM “knows” but they differ in critical ways - much like long-term vs. short-term memory in the brain. Memory is the agent's long-term recall - the knowledge it retains across sessions about you, your organization, your data, and your past projects. But to use any of that knowledge in answering a specific prompt or agentic process, you must bring the necessary information into the LLM’s short-term memory for use in a specific session - Context is that short-term memory. We’ll learn how to identify what you need in Context to support a particular workflow.
    • Data connections 101
      Every agent platform provides built-in connections to a range of apps / data sources. You’ll learn how to safely use these connectors in your agents.
    • Data connections 201
      If a platform lacks a built-in data connector, the MCP industry standard steps in. You’ll explore using MCP servers to connect to data sources that don’t have built-in connectors. And we’ll show how you can go beyond MCP with 3rd party integration products like Zapier.
    • Data Protection & Privacy
      Granting AI tools access to operational applications can introduce vulnerabilities when implemented without care. We need to protect against privacy breaches or unintentional actions. Master the concept of minimum necessary access by assigning tools only required permissions. You’ll learn to set guidelines for safeguarding sensitive client information, securing credential keys, and evaluating vendor terms so your custom processes never train public networks. You will create a straightforward security framework designed to shield your business.
    • Validating and auditing your agents
      Your work doesn't end when an agent completes its initial run. For small teams, an unchecked agent can disrupt client communications, mismanage prospective leads, or waste operational budget. Learn how to integrate simple validation into your development process and routine operations. We'll practice testing agents against edge cases and executing sanity checks prior to handing over autonomy. You'll also set up basic monitoring routines and human-in-the-loop checkpoints to track performance, review tool usage, and keep your agents aligned with your standards over time.
    • Agentic Tools
      The general purpose platforms for creating agents are evolving at incredible speed. Our goal in this program is not to help you master any single platform. It’s to help you understand, as a non-technical professional, how agentic AI works and how you can use it to take operational tasks off your plate. We'll use two or three platforms to give you the hands-on experience you need, and we’ll help you understand how the common concepts apply across them. Learn more about the AI tools we intend to use.

    How will you learn the material?

    Live Online Instruction
    Meet via Zoom with an instructor and your classmates two evenings a week, with an instructor to answer questions in the moment.
    Hands-on Building
    Our classes are an active learning environment, not a place for passive listening and note-taking. You'll build agents for a wide range of problems, from customer interaction to marketing, sales, finance, and operations.
    Exercises + Your Own Workflows
    Work on exercises and examples we provide, plus examples from your own organization and workflows. Class exercises will be based on real-world scenarios in a simulated company environment. Homework exercises will often offer a choice of using the simulated company environment or allowing you to practice using your own organization’s workflows and data, if appropriate.
    Support outside of class hours
    The class will have a Slack channel to support collaboration outside of class hours. This channel provides a way for students to interact with each other and with instructors, get questions answered, highlight things they are struggling with or would like to go deeper into during a class session, etc. We all can learn more and better with the help of others (and when helping others) and this is one way to facilitate continuous learning. There will also be designated “office hours” when instructors will be available to answer questions.

    • Schedule

      Tuesdays and Thursdays 6 PM - 8:30 PM CT

    • Location

      This class is live online (i.e. synchronous).

    • Dates

      See Schedule Below

    • Tuition

      $1150

    Requirements

    Student requirements

    • At least 18 years of age.

    • Live in the Eastern, Central, or Mountain time zones in the U.S.

    • No programming or IT experience needed. This class is designed for non-technical professionals.

    • You use a computer and the internet regularly and are proficient in navigating both. You’ve interacted with an AI chatbot to ask questions.

    Hardware/Software requirements

    Upcoming Sessions

    Program Type Format Dates Tuition
    AI Fluency: From Prompting to Agent Orchestration Evening(10 Weeks) Live Online January 19 - March 25, 2027 $1150 Register


    Payment Details

    You will receive an invoice from Freshbooks within 2 business days after you register. Payment is due upon receipt of the invoice to secure your seat and may be paid via ACH, check, debit card, or credit card.

    Submitting the registration form does not directly reserve a seat in the class. You will receive an invoice via email from Freshbooks within 2 business days after you register. Payment is due upon receipt of your invoice and will hold your seat, provided we're not on a waitlist. Should all seats be reserved at the time of your payment, you will be added to the waitlist. If there are cancellations, you will be invited to the class based on the date your payment was made.

    Refund Information

    If you cancel your enrollment in this class prior to the first class session you are entitled to a refund of 100% of the fee paid to Nashville Software School. If you fail to attend the first and any subsequent class sessions, you are entitled to a refund of 100% of the fee paid minus a $100 administrative fee. If you attend one or more classes and withdraw from the program before attending the third class session, you are entitled to a refund of 90% of fees paid or if you withdraw before attending the fifth class session you are entitled to a refund of 80% of of fees paid (in either case a $100 administrative fee will be charged). No refunds are available after the start of the fifth class session. Note: this policy is in compliance with Tennessee postsecondary education regulations rule 1540-01-02-.17.

    Have questions?
    Email Us