AI Product Engineer. 20 years shipping production systems across fintech, identity, crypto, and high-reliability platforms -- now applying that discipline to agents, harnesses, and applied AI.
Think NCO, not general -- in the trenches with the team, shipping under real constraints.
making-minds.ai -- Research -- Substack -- LinkedIn
| mnemos | Biomimetic memory for coding agents | pip install mnemos-memory[mcp] |
| cartograph | Map the repo before you burn context | npm i -g @anthony-maio/cartograph |
| pulldown | HTTP-first web content for LLM pipelines | pip install "pulldown[mcp]" |
| halobot | MCP Discord channel for any agent | npm i -g halobot |
| slipcore | 60-85% token reduction for multi-agent comms | pip install slipcore |
| synthesis | Federated skill ecosystem for AI self-extension | pip install synthesis-mcp |
Sites: mnemos -- cartograph -- pulldown -- halobot -- synthesis
Scalable AI Oversight -- where verification breaks down and how ensemble fixes help. Verification Swarm -- CMED -- HDCS -- argos-swarm
Multi-Agent Coordination -- efficient, safe protocols for agent swarms. Slipstream (82% token reduction) -- Covert Channel Prevention -- Ollama model
Cognitive Architectures -- persistent memory, coherence, safe self-extension. Coherence-Seeking -- Continuity Core -- Self-Directed Knowledge -- Synthesis -- CoDA-GQA-L (9.5x KV compression)
AI Safety & Alignment -- structural failure modes: sycophancy, hallucination, introspection. Epistemic Dissonance -- Scaffolded Introspection
Applied AI -- Concrete Intelligence: deployment guide for heavy industry (public domain book).
Long-form on Substack. Recent picks:
- Stop Letting Your Users Pick the Model -- AI product engineering for the post-frontier era
- The Agentic Org Chart -- why functional departments are structurally wrong for what comes next
- Ten Agents Destroyed Production and Everyone is Strangely OK With It
- Your Agent Has Amnesia -- announcing mnemos
- The Recursive Developer -- how agentic masters justify $2,000+/mo in coding assistants
- The 80/20 Lie: Why 80% of Agentic AI Work Isn't AI
Also: The Checkpoint (weekly), Medium, Hugging Face blog.
ORCID -- Google Scholar -- ResearchGate -- Hugging Face -- LinkedIn -- X