Generative AI Innovation: From Idea to Working Prototype

The problem this course solves

Nearly every organization has a long list of possible generative AI use cases, fed by industry benchmarks, articles and other people's demos. The problem is not a lack of ideas: it is the lack of a method to separate noise from real value, and an accessible way to take the best idea to a working prototype that proves it.

This creates a familiar pattern:

• Many AI ideas, with no clear criteria to decide which to prioritize.
• Generic use cases ("automate with AI") that never connect to a specific business problem.
• Prototypes that depend on a scarce development team, with weeks of waiting just to validate whether the idea makes sense.
• Proofs of concept that impress in a demo but have no business case or scaling plan behind them.

This course addresses exactly that bottleneck: participants go from identifying and prioritizing real use cases to building, with their own hands and in natural language, a functional prototype that demonstrates tangible business value.

Who it is for

No programming knowledge required. It is designed for the people who lead AI innovation, not necessarily those who code it.

• Leaders and innovators driving AI transformation in their organization.
• Managers and directors of innovation, digital transformation or technology.
• Product Managers and Product Owners.
• Strategy consultants in innovation and technology.
• Entrepreneurs building AI products.
• Engineers (systems, industrial, software) who lead AI projects without being the ones who program.

Prerequisites

• Experience leading projects or teams.
• Basic knowledge of generative AI (using ChatGPT or a similar tool).
• Strategic business or technology perspective.
• Not required: advanced programming or formal computer science training.

What participants take away

• An AI adoption strategy with roles and responsibilities mapped for their organization.
• A repository of at least 10 identified, evaluated and prioritized use cases, with a justified top 3.
• A product design document for the selected use case: user stories, architecture, flows and success criteria.
• Their own working PoC, built in natural language and ready to show to stakeholders.
• The business case (ROI) and the executive presentation for that PoC.
• An implementation and scaling roadmap.

Learning objectives

By the end of the course, participants will be able to:

1. Apply organizational generative AI adoption strategies, identifying key roles (Executive Sponsor, Senior Management, Innovators) and their responsibilities.
2. Identify, evaluate and prioritize generative AI use cases using structured frameworks for feasibility, impact and resources.
3. Apply product development and creative thinking methodologies to design prototypes that solve real business problems.
4. Implement functional PoCs using natural-language tools (ChatGPT, Google AI Studio, Cursor), without advanced programming knowledge.
5. Develop adoption and scaling roadmaps that account for organizational, technical and governance factors.
6. Present value propositions and business cases for generative AI projects to executive and technical stakeholders.

Course content

1. AI foundations and organizational adoption

   Overview of generative AI and the course tooling. Introduces the key roles in strategic adoption (Executive Sponsor, Senior Management, Innovators) and the strategies for scaling from pilot to production. Closes with the role map and a draft adoption strategy for the participant's organization.

2. Identifying, evaluating and prioritizing use cases

   A methodology for spotting opportunities by process and functional area, plus an evaluation framework covering technical feasibility, business impact, required resources and risks. Includes building a prioritization matrix (impact vs. effort) and the value proposition for the selected case. Closes with at least 10 identified use cases and a justified top 3.

3. Prototype design with product methodology

   Applies design thinking and creative thinking to the design of an AI PoC: MVP definition, user stories, flow architecture, data sources and success criteria. Translates all of it into natural-language technical specifications ready to build. Closes with the complete PoC design for the prioritized use case.

4. Building the PoC in natural language

   Step-by-step construction of the PoC using ChatGPT, Google AI Studio and Cursor as natural-language development tools. Includes technical and business documentation, preparation of the executive presentation (demo, business case, ROI) and the scaling plan. Closes with the working PoC implemented and presented to the group.

Methodology

Duration and format

• Duration: 24 hours.
• Suggested format: live virtual sessions — for example 12 sessions of 2 hours or 4 sessions of 6 hours.
• Delivery: virtual, by videoconference.
• Tools: ChatGPT, Google AI Studio and Cursor — free tier as a baseline, no paid subscriptions required.

About the instructor

Jorge Domínguez is a systems engineer with more than 18 years of experience in software development, solution architecture and technology innovation. He holds an MBA from Universität Leipzig, Germany, focused on innovation and technology adoption, and is a graduate of the MIT xPRO executive program in AI product and service design and development (2025). He is certified in Generative AI Fundamentals by Databricks and in Teaching the AI Fluency Framework by Anthropic.

He has led curriculum design and delivery of specialized AI training at Universidad de los Andes, and has trained more than 60 companies and institutions — including Bancolombia, Comfandi and Universidad de los Andes — across the three levels of organizational adoption. He specializes in strategic AI adoption, generative AI application development and AI-assisted software development.

Depending on program scope, the course can be delivered with support from co-instructors specialized in technical or data-intensive cases.

Next step

If mandate and governance are not defined yet, start with AI Governance — 4Ds Framework. If the prototype needs to become a production system, the AI Adoption & Governance Sprint covers architecture, controls and the implementation roadmap.