Sequence Markets: sub‑20ms institutional SDK
Sequence Markets said its WASM‑based SDK delivers ~1μs callbacks and sub‑20ms end‑to‑end latency for institutional trading, positioning WASM as a low‑overhead path for fast AI‑driven strategies. If verified, that kind of callback determinism changes where you can safely run strategy logic without moving entirely onto FPGA or NIC firmware. (x.com)
Sequence’s sequence-algo-sdk repository (Rust → WASM, “zero dependencies”) went public on GitHub two weeks ago and includes an Algo trait plus order-book types and action buffers. (github.com) The SDK is published as sequence-algo-sdk 0.1.0 on crates.io, with package metadata and release tracking visible on Libraries.io. (libraries.io) Sequence’s docs provide example scaffolds that compile a Rust project to the WASM target and show a CLI + Terminal “Algo Studio” flow for uploading.wasm binaries. (docs.sequencemkts.com) Their backtesting engine replays historical L2 book snapshots and trade ticks through the user’s WASM algo, producing deterministic fills, PnL and performance metrics given the same seed and input data. (docs.sequencemkts.com) The public API spec encodes prices as fixed-point 1e9, quantities as 1e8, and records timestamps as nanoseconds since the Unix epoch. (api.sequencemkts.com) Architecture documentation shows an “Algo Manager” that stores and deploys WASM binaries to edge nodes and a Market Data Aggregator that consolidates feeds from those edges into unified books. (docs.sequencemkts.com) Sequence Markets markets itself as an execution layer backed by “$15B+ liquidity” and lists Y Combinator Winter 2026 (W26) backing with founders Peter Bai and Muhammad Awan named on the company/YC pages. (sequencemkts.com)
