Things I built to understand.

These are working systems, experiments, and personal tools from my current chapter.

Some are public. Some are private. Some are small on purpose.

I include them because they show how I now learn: by building with AI, connecting real data, testing ideas directly, and letting reality sharpen the thinking.

Primary systems

Personal capital allocation

Private Investment OSprivate system, used personally

What it is

A private investing and decision-support system connected to broker and market data (read-only). It maps portfolio exposure across conviction sleeves, versioned sleeve sensors, on-chain adoption, macro and regime gates, relative-strength leadership, and a structured review meeting with decision receipts. This system is built for pre-committed discipline, not automated trading.

Why I built it

I built it to turn investing from scattered opinions into a system I can review, improve, and hold accountable over time.

What it explores

AI-assisted capital allocation, personal decision infrastructure, and what becomes possible when one person can build institutional-style review tooling for their own use.

Investment intelligence / AI allocation

Crypto That Matterslive research system

What it is

A research and portfolio-allocation system that compares mechanical rules and independent AI allocators across crypto, stocks, macro, and Bittensor. It keeps price leadership, fundamentals, and value separate, records decisions forward-only, and tracks paper portfolios against relevant benchmarks.

Why I built it

I wanted to see whether AI could do more than explain markets after the fact — whether it could make explicit allocation decisions, preserve the evidence and uncertainty behind them, and then be held accountable for what happened next.

What it explores

AI as an accountable allocator: how agents combine different kinds of evidence, maintain decision memory, disagree with each other, handle uncertainty, and whether their judgment adds anything beyond mechanical rules and passive exposure.

Open site

Bittensor

dTAO AnalyticsBittensor research system

What it is

The Bittensor research system that grew into Crypto That Matters. It tracks subnet markets from my own Subtensor node and combines market structure, fundamentals, flows, emissions, risk, signals, benchmarks, and independent AI portfolio decisions.

Why I built it

Subnet investing was noisy and difficult to evaluate. I wanted the underlying network data, investment evidence, and portfolio decisions in one system rather than relying on narrative and social signal.

What it explores

How decentralized intelligence can be measured as an investable network — and whether AI allocators can turn noisy, fast-changing subnet data into disciplined decisions.

Explore Bittensor

Private AI / confidential TEE

Conclaveworking system

What it is

Private-first chat through confidential TEE inference. Optional Web Research stays visible and separate from private history. Browser-verified E2EE on supported models.

Why I built it

I wanted an AI chat where the boundary is the product—not a privacy policy bolted onto a generic stack.

What it explores

What verifiable private inference looks like when attestation, encryption, and user-visible research boundaries have to hold together.

Open site

Explorations

Semantic discovery

GitHub Resonanceexperiment

What it is

Discovery for builders, repositories, and projects by similarity and technical signal—not keywords and follower counts alone.

Why I built it

Technical hiring and curiosity deserved a deeper map than search boxes usually offer.

What it explores

What relevance means in an open-source graph when embeddings sit next to human intent.

Open site

Private reflection

Echopersonal tool

What it is

A local-first journal where new reflections connect to older ones by meaning, on your device—no cloud account, no training on your words.

Why I built it

Inner life should not be monetized or mined, but it should still remember itself and resurface what matters.

What it explores

Private reflection and learning over time: attention, pattern, and memory without a server in the middle.

Open site

Reflective reading

Gospel Lensretrieval experiment

What it is

A narrow reflective interface: your words meet gospel passages and commentary with calm typography and retrieval you can trust.

Why I built it

Serious texts deserve better than generic chat speed—retrieval, context, and restraint matter here.

What it explores

How RAG and interpretation behave when the source is sacred to someone, not just indexed.

Open site

Resonance / Solana

MuRPworking experiment

What it is

A working Solana-side experiment: on-chain universes, private resonance modes, token-gated spaces, and social discovery without a default feed.

Why I built it

I wanted a constrained place to ask how portable identity and quieter discovery could coexist.

What it explores

Privacy, ownership, and graph dynamics when you refuse both walled gardens and full exposure.

Open site

Physical product

Wooden fly boxes

Landgrenphysical product

What it is

A side-project webshop for handcrafted wooden fly boxes in Skåne, built with my brother.

What's in the shop

Handcrafted beech fly boxes with brass fittings and pocket layouts; optional engraved lids.

Open site

Playful experiment

AI music

Botifyplayful experiment

What it is

A small experiment in AI-generated music and presentation—lightweight, playful, not pretending to be a product roadmap.

Why I built it

Some questions are best answered with a sandbox: taste, tooling, and how far “good enough” media can go.

What it explores

Creative interfaces when the model is the instrument and the joke is part of the point.

Open site

Before this chapter, I spent decades inside critical infrastructure.

That shaped how I think about systems, risk, trust, and what has to work.

I do not want to return to that world as a résumé identity. It is simply the foundation under the work now.

If it resonates, reach out.

Reach out