Deep Dive on AI Context Parameters and Controls
This is a presentation I did for the full stack actuarial conference. It represents the same process I use to enable investor analysis with AI.
The presentation below is intended to be very novice friendly, up front. It is LLM model agnostic and (if I’m honest) use-case agnostic. What you will find in the first half is a demo of the hidden levers of the language model. In the second half, you’ll learn how to create your own agents with my 3rd version agent generation workflow (downloadable via the link below).
The last segment is a demo of creation of a sophisticated calculation process and demo of a formal development workflow. Suffice to say, development of the most complicated financial calculation and forecasts are entirely possible with nothing more that the tools provided in the link below and your institutional knowledge of the task. A non actuary’s take-away could be the questioning of your own skepticism toward the potential precision of the language model (given precise enough inputs and instructions).
Goodies: https://github.com/5thgenfinance/full-stack-actuarial