NYC School of Data 2024
How citymeetings.nyc Uses AI to Navigate and Research NYC Government Meetings
![NYC School of Data 2024](/static/4e784bea346b0bceb0d35496e0b7fb8d5e638947/images/nyc-school-of-data-2024-cartoon.png)
Slides
I've published my annotated slides on my blog.
The talk recording is not yet available. I'll post it here when it is.
Resources
- Prompts used in citymeetings.nyc (context in slides)
- My speaker identification prompt (8k tokens of plain old English)
- My chapter extraction prompts (in code, but still legible -- uses Instructor)
- Blog posts
- Why you should regularly and systematically evaluate your LLM results (vikramoberoi.com)
- A UX-centric approach to navigating city council hearings with LLMs (vikramoberoi.com)
- Think of language models as a "calculator for words" (simonwillison.net)
- Misconceptions about large language models (simonwillison.net)
- Guides
- OpenAI's prompt engineering guide (openai.com)
- Tools & libraries I use
- TypingMind: the chatbot I use, easy to use across LLMs.
- Instructor: how I reliably get structured data from LLMs.
Follow me
- Threads
- Bluesky
- I write at vikramoberoi.com
Workshops and consulting
Level up your organization's use of large language models (LLMs)
I'm available for workshops for both non-developer and developer audiences.
Contact me at workshops@vikramoberoi.com.
Build your product and hire your founding team
I help early-stage non-technical founders build their product and hire their founding teams via my boutique consulting firm, BaxterHQ.
Contact me at vikram@baxterhq.com.