I build tools like SmartEdgar, MCP Agent Mail, and Ultimate Bug Scanner, plus an ecosystem of open-source infrastructure for AI agents and serious research workflows.
And I consult to PE funds and hedge funds about AI automation after a decade as an analyst in the long/short equity hedge fund space.
I also founded and run Lumera Network (formerly Pastel Network), a Cosmos-based L1 for storage, AI, and interoperability.
Generalist long/short across platforms and partnerships.
Studying neural nets since pre-Transformer days.
Agent tooling, infra, research notes, and experiments.
Analysts, founders, researchers, and engineers.
Loading visualization...
Snapshot
Three overlapping threads: infrastructure, tools for agents, and research that actually moves markets.
Agent tooling
Tools that let analysts and LLMs cooperate instead of fighting PDFs, Git diffs, and static analyzers.
Markets & research
Long-form work that connects model internals and infrastructure economics all the way back to cash flows and valuations.
Infrastructure
A Cosmos-based L1 for durable storage and AI verification, aimed at the world where agents talk to chains as fluently as to APIs.
Projects
A comprehensive collection of the tools, protocols, and experiments I’m building.
Writing
A mix of public writing and GitHub-native research artifacts.
A deep technical dive into replacing static Transformer weights with metabolic, fatigue-prone, and evolving biological analogs. Explores presynaptic fatigue, Hebbian fast-weights, and structural plasticity.
What happens when GPT-5 acts as Principal Investigator? An exploration of 11 exotic architectures including Matrix Exponential Gauge Learning, Tropical Geometry, and p-adic Attention.
An interactive deep dive into the Covariance Matrix Adaptation Evolution Strategy. Explains how it adapts to high-dimensional, non-convex landscapes where gradient descent fails.
A 12,000-word deep dive into how AI economics, models like DeepSeek, and GPU supply can collide with valuation narratives. Explores the potential reflexivity of the AI capex cycle.
Uses a forgotten historical prediction—the wave theory of light and the Poisson Spot—as a lens on evidence, updating priors, and how we reason under massive uncertainty.
An insider's look at how 'smart' risk models like Barra often unknowingly distort incentives, encourage crowding, and create hidden systemic risks within multi-manager platforms.
Proposes treating LLM context as a 'save state'. By allowing models to compress and manipulate their own internal thought states, we unlock capabilities like reasoning backtracking and reinforcement learning.
Reconstructs massive fraud patterns in the PPP loan program. Shows how network analysis and simple heuristics could have caught billions in theft that standard banking checks missed.
Demonstrates that collective intelligence outperforms individual genius. A framework where multiple LLMs iteratively refine coding solutions via peer review to break out of local optima.
How to use AI as a tireless patient advocate. Describes a method for synthesizing scattered medical records into a unified dossier to catch missed diagnoses and dangerous drug interactions.
Traces the history of 'jailbreaking' from simple commands to recursive meta-prompts. Argues that internal guardrails always fail and proposes external 'inoculation' strategies.
Advocates for 'separating cognition' when using agents. Instead of asking for code directly, use agents to generate granular plans first, spreading the reasoning load across more tokens.
Argues that internal alignment (Asimov's laws) is a losing battle. Proposes an external 'criminal justice system' for AI: helper models that monitor and police the primary model's output.
The story of the self-taught genius who invented linear algebra (the wedge product) decades before it was understood. A lesson on how radical abstractions are often rejected by the establishment.
Why pipelines aren't enough. Details the 'non-linear exploration model' architecture needed to extract deep, hierarchical structure from complex documents without hallucination.
A technical retrospective on using FastAPI, SQLModel, and Pydantic to build a heavy-duty async backend. Covers Whisper integration and optimizing for concurrency.
A practical guide to decomposing tax returns into context-sized chunks for AI analysis. How to find deductions, spot errors, and out-perform human accountants on complex personal returns.
Standard correlation misses non-linear relationships. This guide explains Hoeffding's D, a powerful non-parametric measure that detects complex dependencies where Pearson and Spearman fail.
A visual, Pythonic implementation of Leslie Lamport's classic solution to the mutual exclusion problem. Illustrates how fair concurrency can be achieved without atomic hardware primitives.
Explains the architecture behind the open-source blogging engine. How to treat GitHub as a headless CMS, using static generation for speed and Git for version control.
Background
A condensed view of the path that got me here.
Balyasny Asset Management L.P.
Generalist long/short equity role at a multi-manager platform, covering both longs and shorts across sectors.
Lumera Network (formerly Pastel Network)
Building a Cosmos-based L1 focused on long-term storage, AI verification, and cross-chain interoperability.
Millennium Management
Long/short equity analyst focused on catalyst-driven situations and idiosyncratic risk.
Dayah Capital
Fundamental research across multiple sectors, with a heavy focus on complex special situations.
Dasoma Capital
Long/short generalist investing in structurally mispriced businesses.
EigenValue Partners
Partner-level role at a long/short fund, leading research and portfolio construction.
Scoggin Capital Management, LP II
Researching long and short ideas across sectors and capital structures.
Tyndall Management
First buyside role, covering a wide array of industries as a generalist analyst.
Reed College
Formal math training plus a lot of time thinking about how abstractions map back to reality.
Threads
I write a lot more informally on X. Here are a few good entry points.
Loading visualization...