AI Dayton Coffee Talk: Understanding and Managing the AI Context Window: How to Get Consistent Results

​Sponsor: Aeroseal

​Speaker: Michael Martin, Founder and Chief Product Officer at Digital2DNA

​Speaker Bio: Michael Martin is Founder and Chief Product Officer at Digital2DNA, based in Bellbrook, Ohio. He has spent more than 20 years at the intersection of technology, healthcare, and product delivery, and has been applying AI to real production problems since 2020. He holds an AI strategy certification from MIT Sloan and was named CIO of the Year by the Cincinnati Business Courier in 2019.

​Event Description: If you've used ChatGPT, Claude, Gemini, or Copilot for real work, you've hit the same wall: results that drift, "the AI forgot what we were doing," or answers that contradict yesterday's session. The culprit is almost always the context window, not the model. In this session we'll open the hood on how context windows actually work, compare how the major tools handle them differently, and walk through live examples of where things go wrong and how to fix them. You'll leave with a concrete playbook for prompt structure, when to start fresh, how to anchor with files and instructions, and how teams can standardize patterns that produce repeatable results across people and projects. Aimed at technical folks and the AI curious. No coding required, but engineers will get extra value from the implementation details.

​Directions: Turn at the large Aeroseal sign at 225 Byers Road. Please park in the visitor spots. If those fill up, you can park where you are able, and go through the main door where the lobby is. There will be someone there to point you in the right direction. See you there!