About
I build AI systems where the model reasons and a deterministic engine computes — agentic harnesses, multi-modal pipelines running local models end-to-end, and the audit-grade data layers underneath them. It started with my own books: I’ve traded crypto since 2020, and when I had to do my taxes nothing on the market could reconcile activity spread across many wallets and chains — so I built my own ledger. FTX in 2022 turned it into a loss-quantification tool and pushed me into local LLMs for unstructured evidence; Claude 3 in 2024 put the model in the loop while the deterministic engine keeps the books. That stack now runs as a business, and building it has thrown off products of its own.
Open to AI and full-stack engineering roles — remote, or based in Tel Aviv.
Work
Tech Stack
Projects
Extends /insights, /usage, claude agents and hooks
Agentic workflows: guardrails and devtools for collaborating agents. Claude Code’s docs tell you what to do when you ask; Fairway tells you what to do when you don’t know what to ask, and gives you hooks to encourage patterns and discourage antipatterns. Nine small tools sharing one Go binary, one WAL-mode SQLite store and one MCP server, so anything one tool writes another can read — each extending a command Claude Code already ships: /insights, /usage, claude agents and the tool hooks. Includes a concurrency guard that intercepts wide git operations (reset --hard, force push, checkout .) before one autonomous coding agent clobbers another’s uncommitted work — multi-agent-safe by design, no shared state between sessions, pre-run interception at the hook layer. Pick the ones you want: most are read-only pull commands that need nothing installed, and two wire into your setup with explicit install and uninstall verbs. The binary alone creates no files, opens no ports, writes no config. Same idea as the ledger work one level up: derive the verdict from observable state instead of trusting anyone to remember.
Done-for-you · the messy tail worked by hand
Koinly and CTC handle the easy 90%; Ledger AI handles the hard 10% — the DeFi protocols and decentralised exchanges the automated tools give up on, decoded into statements accountants and tax offices accept. Works that tail by hand, at volume, and quantifies the gap in dollars rather than hiding it. Also includes fiat extensions for Xero and MYOB.
Accounting components that move correctly
A React component library powering Ledger AI’s front-end. Animated, Excel-literate accounting primitives — journal entries expanding to their debit and credit legs, transaction split and merge, reconciliation matching, P&L, balance-sheet and pivot tables, and entry forms with spreadsheet input semantics. The domain constraint is the product: components take double-entry-shaped props and animate only the transitions the ledger permits, so a split visibly conserves its parent total. It composes over a headless engine rather than wrapping a data grid.
Writing
See allBuilding in Public
I write about building crypto and financial-data tools — and what I learn using AI to ship faster.