Jeffrey Emanuel: Agentic Coding Tooling, AI Infrastructure & Markets

A Claude Code / Codex skill that handles the expensive intake portion of estate planning: four steps to install, one weekend of work, roughly $120, and a structured document package for attorney review.

Two paired Claude Code / Codex skills that move a company from Slack to self-hosted Mattermost — fail-closed readiness gate, named rollback owner, ~99% lower ongoing cost.

Why giving AI models too many constraints degrades output quality.

This project emerged from a simple question I posed to an advanced AI: *"How can we use matrix exponentials to improve attention?"

*Written on 3/22/2025 by Jeffrey Emanuel* !

**By Jeffrey Emanuel (and various collaborators of the electronic persuasion)** *Written on April 1st, 2025* !

!

Standard Large Language Models are "frozen crystals."

The Bakery Algorithm was invented by Leslie Lamport in 1974 (Original paper at [https://lamport.azurewebsites.net/pubs/bakery.pdf] ).

The Birth of an Industry Standard In the mid-1970s, a Berkeley economics professor named Barr Rosenberg fundamentally changed how institutional investors think about risk.

How should you approach complex changes to already complex software projects when using agentic coding tools such as Claude Code or Cursor?

A deep dive into RaptorQ (RFC 6330), the fountain code that turns any file into an infinite stream of interchangeable packets with 0.02% overhead.

Introduction and Background I recently revisited an [article](https://simonwillison.net/2022/Sep/16/prompt-injection-solutions/) by Simon Willison from September of 2022 about prompt injection attacks against large language models (LLMs).

Introduction Healthcare in America is deeply flawed, a fact painfully evident to anyone managing complex or chronic medical conditions.

Step 1: Prepare Your Tax Return PDFs Get complete PDFs of your last 3 years worth of complete tax returns.

Introduction In the crowded space of LLM applications (referred to in a denigrating sense as "wrappers"), most tools follow a predictable pattern: feed text to an LLM and receive a straightforward response.

My Favorite Black‑Box Optimizer: CMA‑ES **by Jeffrey Emanuel** If you live anywhere near modern machine learning, “optimization” almost automatically means *gradients*.

The Paycheck Protection Program (PPP) was a cornerstone of U.S. economic relief during the COVID-19 crisis, disbursing nearly $800 billion to small businesses.

As someone who spent ~10 years working as a generalist investment analyst at various long/short hedge funds (including stints at Millennium and Balyasny), while also being something of a math and computer nerd who has been studying deep learning since 2010

I've always been impressed when people can truly predict the future.

I stumbled onto the idea for [this app](https://youtubetranscriptoptimizer.com/) by seeing how much time my wife, who is a YouTube [creator](https://www.youtube.com/user/GracieTerzian) who makes videos about learning music theory, spent preparing addition…

Introducing Our New Next.js Blogging System with Tailwind CSS A Sleek and Efficient Way to Share Your Thoughts !

I believe that it's basically impossible to make a single inherently safe AI in the form of an LLM or agent-based system.

I periodically think about [Hermann Grassmann](https://en.wikipedia.org/wiki/Hermann_Grassmann) and get sucked into reading about his life and work.

I recently needed a blog for my recent Next.js app, [YouTube Transcript Optimizer](https://youtubetranscriptoptimizer.com/), and wanted something that would look really nice and that could be integrated into my existing Next.js app to keep deployment simp…