ONLINE COURSE
by KEDAR UNDALE

From Parametric Patterns to GenAI Animation

Built for designers who want AI in the workflow without losing your vision.

From Prompting to Negotiation

This course is built on three principles: 
future proof learning, maintaining design intent and cost effectiveness. 

This is not a fast AI workflow. Each output will be built with intention and will take time. If your goal is to input a prompt and produce an image, this course does not follow that approach. Instead, it focuses on maintaining control over your design, understanding how the tools we use interpret our data, and translating geometry into pixel information in order to build a consistent image catalogue. In the course, we will dive deeper into the logic behind this process. 

Within parametric design, it covers data trees, extracting parametric frames from Grasshopper, and preparing structured data so that AI systems can interpret and render it correctly. We will also focus on maintaining consistency in pixel information across different tools so all our outputs remain aligned.

A key idea in the course will be “negotiation”. You will define the design and test how AI interprets it through probability, instead of asking AI to simply generate design answers. We will use AI as a data and rendering system, and not as a co-designer. We will focus on using the capabilities of AI while maintaining authorship.
In this course, we will shift the focus from prompting repeatedly to building structured systems and master prompts. 

Instead of merely writing prompts one by one, you will create a framework where your input images and systematic rules will generate consistent outputs. This approach moves toward building parameters that AI operates within rather than describing outcomes manually.

The workflow will extend into creating detailed zoomed-in images that will be part of the same system. These are automated but controlled, adding variation and spatial depth while maintaining consistency. The intention is to create outputs that read across scales and feel structured rather than isolated.

At the end of the course, you will not just replicate the workflows shown but understand in-depth how to build your own systems. You will learn how to translate your design intent into a format and language that AI can work with and, guide it towards outputs that stay consistent with your original vision.

A quick glimpse into the course:

Write your awesome label here.

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

  • Understand how to build your own systems rather than just replicate ready-made workflows.
  • Translate your design intent into a format AI can work with.
  • Guide AI toward outputs that stay consistent with your approach.
  • Gain intermediate level proficiency in AI-driven processes for parametric design.

✓  What's included in the course?

Go from Rhino to Render.

RHINO 8
Grasshopper
COMFYUI
ADOBE PREMIERE PRO
PRACTICAL EXPERIENCE DISTILLED INTO 

HOURS OF EXPERT AI LESSONS.
WHEN IT COMES TO AI, LESS IS MORE.

ROBUST TOOLS ARE ENOUGH TO GET THE JOB DONE.
MASTER THE CONCEPTS OF AI WITH

QUIZZES & ASSIGNMENTS.
AUTHENTICATE YOUR AI LEARNING JOURNEY WITH

CERTIFICATE OF COMPLETION.

We will focus on 4 Fundamental Processes.

01

Developing parametric patterns inside Grasshopper3D.




02

Translating those patterns for AI, then running them in ComfyUI/ComfyCloud via API nodes - no GPU limits.

03

Converting the parametric pattern and render into a smooth animation using API workflows and Adobe Premiere Pro.

04

Making the design come alive with intimate, ground-level shots seen through human or child eyes, rendered in ComfyUI.

How much will the workflows cost?

  • You do not need any subscription fees to any tools used in the course.
  • The only cost is the API usage, which is typically around 2 to 3 USD per workflow and includes the main render, animation, and zoomed images.
  • This results in around 10 professional stills and a video output from 3-4 workflows.
  • The focus is not just on cost but the ability to generate consistent, scalable, and controlled results from your own data.

Your Course Instructor

Kedar Undale is an architect, educator and designer working at the intersection of architecture, parametric design, and generative AI. Through his independent practice, he collaborates with architects and designers seeking to realize complex, form-driven designs. His research in Generative AI is primarily focused on methodologies that speed up such complex design processes. 

Kedar holds a Master’s degree from IAAC Barcelona and has worked as a Computational Designer at the Toronto based architecture studio PARTISANS. Currently Kedar is the AI and Visualization Lead at Bienenstock Natural Playgrounds, Hamilton, where he combines parametric design thinking with generative AI to develop visualization workflows for playground design.

Over the past seven years, Kedar has taught global workshops on ‘Parametric Thinking’ at institutions like DigitalFUTURES, InterAccess Canada, POWRPLNT New York, and CEPT India. Since the advent of generative AI, Kedar has delved into creating custom workflows, leading training sessions at organizations and faculty development programs including PAACADEMY, UID Gujarat, and KLS GIT Belagavi. His art, featured in Homegrown Magazine India, leverages computational tools to visualize intangibles like sound, wind, and gravity. 

Course Details

Keep Authorship. Build Systems. Render at Scale.

Empty space, drag to resize
Write your awesome label here.
Empty space, drag to resize

If you are an organization with more than eight members or employees, you may be eligible for discounted rates or customized pricing. Please write to us at courses@kedarundale.com.

Created with