Public Event

In Person

Arize Builders Meetup - NYC - Boosting Claude Code performance with prompt learning

Register for Event
RSVP

Mar

12

Your CLAUDE.md file is more powerful than you think.

In this talk, we'll walk through how we applied Prompt Learning — an RL-inspired prompt optimization technique — to improve Claude Code's performance on SWE-Bench Lite by up to 11%, purely by optimizing the system prompt instructions. No fine-tuning, no new tools, no architecture changes. Just better prompts, driven by real performance data and LLM-as-a-judge feedback.

​We'll cover the full optimization loop (rollouts → LLM evals → meta-prompting), show results for both general coding improvement and repo-specific specialization, and share practical takeaways you can apply to your own coding agent workflows today.

when

Thursday
March 12, 2026
6:00 pm

LoCATION

Betaworks • 29 Little West 12th Street, New York, NY 10014

agenda

6:00 - 6:30 PM | Check-in & Networking
6:30 - 7:30 PM | Presentation & Q&A
7:30 - 8:30 PM | Networking

​Food and drinks will be provided. Space is limited—register soon to secure your spot!

Speakers

No items found.

Sponsors

In Person
–
Public Event

Arize Builders Meetup - NYC - Boosting Claude Code performance with prompt learning

March 12, 2026
6:00 pm

Postponed: New Date TBA

RSVP

About

Your CLAUDE.md file is more powerful than you think.

In this talk, we'll walk through how we applied Prompt Learning — an RL-inspired prompt optimization technique — to improve Claude Code's performance on SWE-Bench Lite by up to 11%, purely by optimizing the system prompt instructions. No fine-tuning, no new tools, no architecture changes. Just better prompts, driven by real performance data and LLM-as-a-judge feedback.

​We'll cover the full optimization loop (rollouts → LLM evals → meta-prompting), show results for both general coding improvement and repo-specific specialization, and share practical takeaways you can apply to your own coding agent workflows today.

When

March 12, 2026
6:00 pm

Where

Betaworks • 29 Little West 12th Street, New York, NY 10014
RSVP

Agenda

6:00 - 6:30 PM | Check-in & Networking
6:30 - 7:30 PM | Presentation & Q&A
7:30 - 8:30 PM | Networking

​Food and drinks will be provided. Space is limited—register soon to secure your spot!

Speakers

No items found.

Sponsors

Proof of Full Vaccination Required

Betaworks requires all visitors to show proof of full vaccination to enter. No exceptions. Two ways that you can show proof:
  • Official vaccine card with at least 2 weeks having passed since the date of the last required dose (1 dose for J&J/AstraZeneca, 2 for Pfizer/Moderna)
  • An active digital pass such as the NY State Excelsior Pass or the CLEAR app

Mar

12

Public Event

In Person

Arize Builders Meetup - NYC - Boosting Claude Code performance with prompt learning

Register for Event
Register Now

Your CLAUDE.md file is more powerful than you think.

In this talk, we'll walk through how we applied Prompt Learning — an RL-inspired prompt optimization technique — to improve Claude Code's performance on SWE-Bench Lite by up to 11%, purely by optimizing the system prompt instructions. No fine-tuning, no new tools, no architecture changes. Just better prompts, driven by real performance data and LLM-as-a-judge feedback.

​We'll cover the full optimization loop (rollouts → LLM evals → meta-prompting), show results for both general coding improvement and repo-specific specialization, and share practical takeaways you can apply to your own coding agent workflows today.

when

Thursday
March 12, 2026
6:00 pm

LoCATION

Betaworks • 29 Little West 12th Street, New York, NY 10014

agenda

6:00 - 6:30 PM | Check-in & Networking
6:30 - 7:30 PM | Presentation & Q&A
7:30 - 8:30 PM | Networking

​Food and drinks will be provided. Space is limited—register soon to secure your spot!

Speakers

No items found.

Sponsors

3/12/2026

Betaworks

Arize Builders Meetup - NYC - Boosting Claude Code performance with prompt learning

Presented by

Your CLAUDE.md file is more powerful than you think.

In this talk, we'll walk through how we applied Prompt Learning — an RL-inspired prompt optimization technique — to improve Claude Code's performance on SWE-Bench Lite by up to 11%, purely by optimizing the system prompt instructions. No fine-tuning, no new tools, no architecture changes. Just better prompts, driven by real performance data and LLM-as-a-judge feedback.

​We'll cover the full optimization loop (rollouts → LLM evals → meta-prompting), show results for both general coding improvement and repo-specific specialization, and share practical takeaways you can apply to your own coding agent workflows today.

speakers

No items found.

Schedule

No items found.

WHEN

Thursday, March 12, 2026
6:00 pm

WHERE

Betaworks
29 Little West 12th Street
New York, NY 10014

Agenda

6:00 - 6:30 PM | Check-in & Networking
6:30 - 7:30 PM | Presentation & Q&A
7:30 - 8:30 PM | Networking

​Food and drinks will be provided. Space is limited—register soon to secure your spot!

Tools
For
Thinking

How New Technologies are Changing How We Create, Share, and Build Knowledge

When

March 12, 2026
6:00 pm

Where

Betaworks • 29 Little West 12th Street, New York, NY 10014
or livestream
Proof of Full Vaccination Required

RENDER is Over, but Camp is just beginning

We had some great speakers and participants turn out for Render, and we all got to participate in very interesting conversations. You can find recordings of all of our sessions below. If you're working on a Tool for Thinking and want to participate in our upcoming accelerator program, you can learn more here.

Building Bicycles for the Mind

Your CLAUDE.md file is more powerful than you think.

In this talk, we'll walk through how we applied Prompt Learning — an RL-inspired prompt optimization technique — to improve Claude Code's performance on SWE-Bench Lite by up to 11%, purely by optimizing the system prompt instructions. No fine-tuning, no new tools, no architecture changes. Just better prompts, driven by real performance data and LLM-as-a-judge feedback.

​We'll cover the full optimization loop (rollouts → LLM evals → meta-prompting), show results for both general coding improvement and repo-specific specialization, and share practical takeaways you can apply to your own coding agent workflows today.

Speakers

Howard

Rheingold

Author
Tools for Thought
@hrheingold
No items found.

Schedule

11:30 AM
Doors Open
12:00 PM
Welcome
with MC
12:10 PM – 12:20 PM
How Do We Define Tools for Thinking and Why Do They Matter?
Join Jerry Michalski and John Borthwick as they talk about their interest in tools for thinking, and what excites them the most about the future of the category.
12:20 PM – 12:50 PM
Inflection Points for Tools for Thinking
What are the key inflection points that will supercharge Tools for Thinking in the near future? How will new technologies, user metaphors, and funding models change how people build these tools? John Borthwick will be discussing how the landscape is changing with the co-founders of Readwise, Daniel Doyon & Tristan Homsi.
12:50 PM – 1:00 PM
Q+A
1:00 PM – 1:15 PM
Demo: Protocol Design for Tools for Thinking
Gordon Brander will be presenting a brief demo on Subconscious and the Noosphere, products he and his team are working on to allow Tools for Thinking to become interoperable and better connected.
1:15 PM – 2:00 PM
Lunch and Networking
2:00 PM – 2:45 PM
The History and Future of Software as Tools for Thinking
Some of the earliest examples of software explored by pioneers like Doug Engelbart, JCR Licklider, Alan Turing, and others were at their core technologies that help magnify, inspect, and spread our ideas. Jerry Michalski will be sitting down (virtually) with Howard Rheingold, author of Tools for Thought, to explore the history and future of "mind-amplifying technology".
2:45 PM – 3:00 PM
Q+A
3:00 PM – 4:00 PM
Tools for Thinking Product Demos
We’ll hear from builders and thinkers deep in the space and get to take a look at what they’re working on. We’ll be checking out Plexus, Re:Collect, Jerry’s Brain, Subconscious and more. 
4:15 PM – 4:45 PM
Leveraging AI and ML in Building New Tools for Thinking
Alice Albrecht and Linus Lee will be sitting down with Chris Pedregal to discuss their work in leveraging AI and Machine Learning for creating new kinds of tools for thinking.
4:45 PM – 5:00 PM
Q+A
5:00 PM – 5:30 PM
Idea Dimensionality and Representing Semantic Meaning
How do ideas - and the human brains that make and hold them - interact?  Get prepared for meta! Esther Dyson and Jerry Michalski, will discuss the idea of how people work together to shape, compare, intertwine and ultimately produce multi-faceted ideas and multi-dimensional idea spaces.  As David Waltz (Thinking Machines) once said, “Words are not in themselves carriers of meaning, but merely pointers to shared understanding.” Watch and lob questions as the two of them try to build and share the idea of how better ideas can be developed through collaboration.
5:30 PM – 5:45 PM
Q+A
5:45 PM – 7:00 PM
Closing and Happy Hour

Proof of Full Vaccination Required

Betaworks requires all visitors to show proof of full vaccination to enter. No exceptions. Two ways that you can show proof:
  • Official vaccine card with at least 2 weeks having passed since the date of the last required dose (1 dose for J&J/AstraZeneca, 2 for Pfizer/Moderna)
  • An active digital pass such as the NY State Excelsior Pass or the CLEAR app
ContactEvent InquiriesPrivacy Policy
© 2026 Betaworks. All rights reserved