Technical Lead

About the project We are a community organisation in remote Australia building sovereign knowledge infrastructure: a decentralised, offline-first platform that lets family groups hold and share significant community and environmental knowledge on their own devices, on their own terms, with no cloud dependency. Knowledge syncs peer-to-peer over local networks; access is governed by community-defined rules enforced in software; an always-on local server archives everything on site. The setting involves genuinely low-connectivity field conditions — offline-first is a hard requirement, not a preference. The platform is built on @comapeo/core (the Hypercore-based engine behind CoMapeo, a production field-data platform designed for offline use by frontline communities), with a custom governance layer on top. The core is working and tested: multi-project archive nodes, LAN peer discovery, a threshold-based attestation system with a full test suite, all pinned and green. Development is community-led with heavy use of AI-assisted coding (Claude Code) inside a disciplined framework: written architecture rules, a phased runsheet with test gates, an append-only build log, and commit-after-green practice. We are not looking for someone to take the project over. We are looking for a technical conscience: an experienced engineer who reviews, hardens, advises, and unblocks — while direction, priorities, and design authority remain with the organisation. What you'll do Code and architecture review (first engagement). Review the existing repository — governance layer, attestation storage, archive-node design, test suite — and deliver a written assessment: correctness, security posture, failure modes, and a prioritised hardening list. The codebase is small and well-documented; this is a bounded first milestone. Security hardening of the governance layer. Our access-control model is threshold-based and fail-closed by design. You'll pressure-test that: authorship verification, spoofing and replay considerations, key management on shared devices, and the sync-boundary behaviour of permissions. Phase support: LAN replication and mobile. Support (not replace) our internal build process through two-device LAN replication testing, then the React Native / Android client built on @comapeo/ipc, following the comapeo-mobile production toolchain. Android native-module and build-chain experience is where we most expect to lean on you. Engineering practice. Stand up lightweight CI (run the test suite on every commit), advise on backup/restore for the local archive server, and review AI-generated code before phase gates are declared passed. Async advisory. Short written responses to technical decisions as they arise; occasional video calls. Required experience Senior-level Node.js (ESM, native modules, testing) — 5+ years or equivalent depth Peer-to-peer or local-first systems: Hypercore/Holepunch ecosystem strongly preferred; otherwise substantial work with libp2p, CRDTs/sync engines (Automerge, Yjs), IPFS, or comparable distributed log/replication systems React Native with native modules on Android, including real build-toolchain debugging (NDK, prebuilds, node-gyp pain) Security review experience on access-control or cryptographic-adjacent code — you can reason about threat models, not just run linters Strong written communication — most of this engagement is asynchronous and documented Strongly preferred Direct experience with CoMapeo/Mapeo or adjacent offline field-data tooling Prior work with community-controlled organisations or under formal data-governance frameworks, or demonstrable understanding of building technology where the community, not the engineer, holds design authority Offline-first deployments in remote/low-connectivity field conditions Experience reviewing/supervising AI-assisted codebases SQLite/Drizzle, Fastify, sodium/libsodium familiarity Data and governance requirements (non-negotiable) The platform will eventually carry sensitive community knowledge governed by strict access protocols. During this engagement you will work only with synthetic test data — no real community data is present in the repository, and none will be shared. Technical recommendations that would weaken community-defined access controls or the fail-closed security posture will be declined regardless of technical elegance; we expect you to flag trade-offs honestly and accept the organisation's decisions on them. All work is contracted work-for-hire; IP and confidentiality terms will be set out in the contract. Portions of the codebase are intended for eventual open-source release and upstream contribution — a plus if that excites you.