About
The AI-native OS for commodity trading
We started with the rubber trade, rewrote CTRM on a modern cloud-native stack, and built AI in as a colleague that drafts but never posts to the books on its own. The goal is singular: the most advanced operating system for commodity trading.
What we're building
xRubber is an AI-native commodity trading and risk management (CTRM) platform. AI isn't a chat box bolted on the side — it's an employee that runs the whole flow, from document intake and trade confirmation to exposure and mark-to-market.
- ✓ Document intake: email / PDF / IM → structured trade, posted only after human review
- ✓ Exposure and mark-to-market, netting reconciliation, credit-line and margin guardrails in one view
- ✓ Rules engine, anomaly center, tasks center, and market-intelligence terminal, out of the box
From one vertical to a platform
We went deep on the rubber trade on purpose, then widened out — adding a commodity means configuring data, not changing code.
- ✓ Today: the leading CTRM for the rubber trade, matching then exceeding legacy practice
- ✓ Next: metals, chemicals, softs and more — one commodity-agnostic model
- ✓ Long term: an AI enterprise operating system spanning ERP-class workflows
Two lines we won't cross
The deeper AI reaches into your books, the harder the boundaries have to be. Private deployment keeps your data in-house; human release keeps AI an assistant, always.
- ✓ AI only drafts, extracts, and warns — a human always reviews and releases financial changes; there is no bypass
- ✓ Local LLM + RAG: private data never leaves your premises, and it runs on a single machine, even offline
- ✓ Decimal-precise money, a complete replayable audit trail, RBAC and separation of duties
Our mission
To build the most advanced operating system for commodity trading — one that fits how the desk actually works. World-class CTRM is the starting line, not the ceiling.
- ✓ A modern cloud-native architecture built to world-class CTRM standards, not a patched-up legacy system
- ✓ Define the rules before we code them: commodity-specific units, precision, and edge cases are first-class
- ✓ Built for trading, operations, finance and compliance teams, global-first