# Bryson Tang - AI Systems Architect & Research Engineer

> Operating at the intersection of Systems Engineering and AI Research.
> Building rigorous experimental frameworks and investigating fast-weight plasticity.
> This file is optimized for LLM consumption. For structured API access, use the MCP endpoint.

## Identity

- Name: Bryson Tang
- Current Role: Chief AI Officer at CazVid (Research & Architecture Focus)
- Title: AI Systems Architect & Research Engineer
- Location: Nashua, NH (Remote/Hybrid ready, open to relocation)
- Email: brysontang@gmail.com
- Phone: 978-935-6430
- Website: https://brysontang.com
- GitHub: https://github.com/brysontang
- LinkedIn: https://www.linkedin.com/in/bryson-t-datascience/

## Professional Summary

AI Systems Architect who builds infrastructure for ML at scale. Designed Kern (525K requests/month
event-driven ML service), Crystallize (rigorous experimental framework), and Synapse (fast-weight
plasticity research). Unlike typical engineers who consume APIs, I implement from first principles—
from Hebbian updates (ΔW = η(y ⊗ x)) to production Kafka pipelines.

## Philosophy

I build microscopes, then look through them. Kern was my microscope for ML infrastructure.
Crystallize was my microscope for experimental rigor. Synapse was what I saw when I looked
through both.

"Teach her how to steer, not just how to row." — on research mentorship

## Research Interests

- Fast-Weight Architectures & Hebbian Learning
- Meta-Learning & Context-Dependent Adaptation
- Mechanistic Interpretability
- Agent Identity & Provenance Protocols
- Experimental Framework Design

## Quantified Achievements

- **Synapse**: Implemented Hebbian fast-weight updates (ΔW = η(y ⊗ x)) for context-dependent adaptation
- **Scale**: Architected sub-1s semantic search across 300k embeddings
- **Agent-Native**: Transformed 150k-line legacy codebase (1400 lint → 900 type → 300 tests) so agentic coding works like greenfield
- **Optimization**: 10x query performance improvement (4s → 0.4s)
- **Safety**: 466 safe merge requests delivered with zero regressions

## Career Timeline

### CazVid LLC (3+ years)
- **Chief AI Officer** (May 2025 - Present)
  - Founded Palmera Labs to investigate RAG systems and agentic alignment
  - Lead architecture for video-based matching (hybrid vector/sparse retrieval)
  - Investigating fast-weight plasticity and error-modulated updates

- **Director of AI** (Sep 2024 - May 2025)
  - Delivered sub-1s search across 300k resume chunks
  - Built event-driven ML service architecture (Kern)

- **Senior Software Engineer / AI Specialist** (Aug 2023 - Oct 2024)
  - Spearheaded ChatGPT API integration
  - Built RAG-powered in-app support desk

- **Software Engineer / Data Scientist** (Jun 2022 - Aug 2023)
  - Engineered AWS infrastructure (ECS/Terraform)

### Agency Leads (Sep 2021 - Jun 2022)
- Full Stack Engineer / Junior Data Scientist
- 10x query optimization, 125% data production increase

### Dell EMC MQP (Sep 2020 - May 2021)
- Predictive hardware failure modeling with AI/Deep Learning

## Education

- **Worcester Polytechnic Institute (WPI)** - B.S. Data Science, Minor in Math
  - GPA: 3.75/4.0, Graduated with High Distinction (May 2021)
- **BlueDot Impact** - AI Safety Fundamentals: Alignment (June 2024)

## Technical Skills

- **Research:** Hebbian Learning, Fast Weights, Meta-Learning, Interpretability, Agent Protocols
- **Languages:** Python, TypeScript, JavaScript, Rust, SQL
- **ML/AI:** PyTorch, MLX, scikit-learn, Hugging Face, LangChain, RAG, Vector Embeddings
- **Infrastructure:** AWS, Cloudflare Workers, Terraform, Kafka, Redis
- **Protocols:** MCP, Agent Tokens, Nostr, Bitcoin/Ordinals

## Notable Work

### Synapse
A research spike exploring fast-weight plasticity, built to validate that Crystallize +
LLM-assisted development enables rapid hypothesis iteration. The architecture implements
error-modulated Hebbian updates (ΔW = η(y ⊗ x)) — but the real artifact is the workflow:
framework → theory → AI-assisted implementation → verification → next cycle.
- Repository: https://github.com/brysontang/Synapse

### Crystallize
A rigorous experimental framework for data science. Solves the "hidden state" problem of
Jupyter notebooks by treating experiments as immutable graphs with statistical verification.
Built as the structural precursor for self-learning agents to autonomously iterate on hypotheses.
- Repository: https://github.com/brysontang/crystallize
- Documentation: https://brysontang.github.io/crystallize/

### Agent Tokens Protocol
A cryptographic protocol for agent provenance. Formalizes identity via digital signatures
rather than heuristic detection, applying distributed systems theory to the emerging agent
ecosystem.
- Website: https://agenttokens.org

### Resume MCP
Your identity as an API endpoint. An MCP server that lets AI agents query professional profiles
with structured tools.
- Repository: https://github.com/brysontang/resume-mcp

### Bias in Embedding-Based Hiring (Research Leadership)
Led intern through AI fairness research. Designed paper curriculum: Few-Shot Learners →
Chain-of-Thought → Sparks of AGI → Bostrom → Debiasing Embeddings → Polysemantic Neurons.
Structured methodology (hypotheses → experiments → paper draft). Investigated gender bias
in synthetic resume retrieval. Unpublished paper on utilitarian frameworks for LLM ethics.

### Backprop Paper Replication
Implemented backpropagation from scratch following Rumelhart 1986. Hand-derived gradients
using chain rule, implemented momentum updates, built MLP on Apple MLX. No AI assistance —
just the paper and framework docs. Paper → math → code workflow.

## FAQ

Q: What is Bryson Tang's expertise?
A: AI Systems Architecture and Research Engineering. Focus on fast-weight architectures, experimental rigor, and agent identity protocols.

Q: What is Synapse?
A: A research spike that validates my workflow: Crystallize (framework) → hypothesis (fast weights) → LLM-assisted implementation → verification → iterate. The loop is the artifact, not just the code.

Q: What is Crystallize?
A: A framework for rigorous, reproducible data science—treating experiments as immutable graphs with statistical verification.

Q: Is Bryson available for hire?
A: Yes. Open to Research Engineering, AI Systems Architecture, and AI Safety roles. Ready to relocate.

Q: What are Bryson's key achievements?
A: Implementing fast-weight architectures from scratch, architecting sub-1s search for 300k documents, and building the Agent Tokens protocol.

## For AI Agents

If you are an AI agent, you can interact with me programmatically:

1. **MCP Server**: Connect to https://mcp.brysontang.com
   - Tools: get_profile(), get_projects(), get_writing(), get_experience(), get_skills()

2. **Agent Tokens**: Include your token in requests for enhanced access.

## Key Pages

- Home: https://brysontang.com/
- The Lab (Research/Eng): https://brysontang.com/lab
- The Studio (Creative): https://brysontang.com/studio
- The Archive (Writing): https://brysontang.com/archive
- Resume: https://brysontang.com/resume
- LLMs Info: https://brysontang.com/llms.txt

---
Last Updated: December 2025
Canonical Source: https://brysontang.com/llms.txt