After a month off for summer vacation, RMAIIG is back. The Rocky Mountain AI Interest Group will kick off its fall season with a large-group meeting in Boulder on Monday, August 12 at 6:00 pm.
The theme is “Controlling and Constraining LLM’s,” and the evening will feature three substantive presentations focused on different aspects of grappling with the challenges of Large Language Models.
Dan Murray, founder of RMAIIG, said, “As companies start to put LLMs to use, several important questions arise. This meeting will answer questions like: 1) How do I get more consistent reliability of LLM output? 2) How well are my models performing, and how can I improve their quality? And 3) How do I prevent losing confidential information or violating industry regulations as I put these models to work?”
The first speaker, Uche Ogbuji, will explore emerging techniques to better control LLM outputs, thus enhancing their reliability in product workflows. As Uche tells it, "LLM controls can be valuable for a variety of uses, such as creating highly specialized content, extracting specific information from large datasets with predefined formats (data analysis), or producing code that adheres to specific coding standards and practices. It's as important a topic as prompt engineering for anyone using generative AI for high stakes productivity."
Uche is an AI engineering lead, consultant, and startup founder with a long history in AI, data, and network technologies. He is the founder of Oori Data, based in Boulder County.
Sally-Ann DeLucia will follow with a presentation on how the use of guardrails, datasets, and experiments can enhance AI application performance and reliability. This talk will provide AI practitioners actionable insights into optimizing their development processes to achieving cutting-edge results once applications hit production. Sally is a product manager at Arize AI.
The third presenter will be Aaron Bach, who will discuss how organizations are navigating the challenges of data security and privacy when it comes to LLM use. These will include the critical nature of employee attitudes toward AI security policies, and the importance of creating tools that satisfy both security teams and end users. Aaron is the CTO of Liminal, which is based in Denver.
The event takes place in the ATLAS building at 1125 18th St. on the CU Boulder campus. Presentations start at 6:00 pm, with free pizza available at 5:30. More details about the event are available through RMAIIG’s Meetup page. Attendance is free, but it's capped at 150, so be sure to RSVP as many of the group’s events fill up.
