MEMORY
Episodic, semantic, graph and working memory—retrieved with purpose, not dumped into context.
HAWTHORN.OSCONNECTING AGENTS
AGENTIC AI ENGINEER // SYSTEM ONLINE
I BUILD BRAINSFOR AIAGENTS.
Memory, governance and self-improving systems for agents that work in the real world.
MEMORY ONLINE⠿
EVAL LOOP 97.4%⠿
GOVERNANCE ACTIVE⠿
✓POLICY GATE
✓EVIDENCE LOG
✓HUMAN REVIEW
SCROLL TO ENTER ↓
SYSTEM STATUSALL AGENTS OPERATIONAL
MEMORY ✦ GOVERNANCE ✦ EVALS ✦ MULTI-AGENT SYSTEMS ✦ RETRIEVAL ✦ EVIDENCE ✦ MEMORY ✦ GOVERNANCE ✦ EVALS ✦
01 / SYSTEM ARCHITECTURE
AGENTS SHOULD
REMEMBER.
PROVE. IMPROVE.
I design the machinery between a model and the real world—the context it receives, the tools it can touch, the evidence it leaves, and the lessons it is allowed to keep.
GOVERNANCE
Policy gates, approval boundaries, provenance, and review-before-apply learning.
EVALUATION
Trace-level evidence, deterministic checks, failure taxonomies, and improvement loops.
ORCHESTRATION
Multi-agent workflows that survive long tasks, retries, handoffs, and real infrastructure.
REAL-WORLD SIGNAL / 2026
142K RECORDS.
ONE HIDDEN PATTERN.
A multi-agent anomaly detection system for a multimillion-dollar dairy company—built to find where subtle quality anomalies were hiding across a vast collection network.
02/03CONFIRMED
600–700+COLLECTION CENTERS2 OF 3AI-FLAGGED CENTERS CONFIRMED IN A SECRET AUDIT
02 / SELECTED SYSTEMS
SOFTWARE WITH
A MEMORY.
Selected open-source systems from my GitHub. Each begins with the same question: what does an agent need around the model to become reliable?
01
LEARNING GOVERNOR
SkillLoop
A local-first self-improvement harness that turns agent traces into governed evaluations, memories, skills, and fine-tuning exports.
↗02AGENT RUNTIME
Governed Agent Architecture
LangGraph orchestration with durable memory, hybrid retrieval, local embeddings, injection defense, and evidence-first execution.
↗03PERSISTENT MEMORY
SuperMem
Four-tier retrieval across SQLite FTS5, a knowledge graph, vectors, and an LLM agent—exposed through MCP.
↗04MEMORY WITHOUT RAG
Recall
An agent-navigated local knowledge base for persistent memory across Claude Desktop, LM Studio, and ChatGPT.
↗05AUTONOMOUS MARKETS
Oculie
A weather-market agent that fuses forecast sources, finds mispriced contracts, and sizes risk using the Kelly criterion.
↗03 / HUMAN IN THE LOOP
ENGINEER.
BUILDER.
STORYTELLER.
AGENTIC AI ENGINEER
Building persistent memory, governed learning, and multi-agent infrastructure for enterprise systems.
xAI · GROK AUDIO
Worked on Hindi and English evaluation and training for multilingual voice intelligence.
FROM DATA TO AUDIT
Built an anomaly-detection system across 142K+ dairy records; two of three randomly audited flagged centers were confirmed.
COMMUNITIES & MEDIA
Helped grow a 100K+ YouTube audience, produced a podcast with 2M+ listens, and hosted large-scale hackathons.
THE FUTURE ISN’T JUST
BIGGER MODELS.
IT’S SMALLER, SHARPER
SYSTEMS AROUND THEM.
04 / OPEN A PORTAL
LET’S BUILD
SOMETHING THAT
REMEMBERS.
Have an agent system that needs memory, governance, evaluation—or a path out of prototype purgatory?
CALENDAR ONLINEIST / WORLDWIDE
OPEN CAL.COM ↗