Built by an engineer who wanted AI speed without giving up security
LogicGrid AI is led by MJ — a Linux and DevOps engineer with more than a decade of experience across infrastructure, containers, security, and distributed systems. The security training came early, and it stuck: a deep, reflexive caution about untrusted systems and sensitive data.
That caution collided with the speed of generative AI. The collision became LogicGrid.
The origin
While building a WhatsApp platform for families and communities back home in Nigeria, AI coding tools compressed weeks of work into hours. But the speed created a new problem. Logs, webhook payloads, config snippets — and occasionally values that should never have left a terminal — were getting moved by hand into AI chats.
The same engineer who had spent years telling people not to expose credentials was doing the AI equivalent. And the workflow itself was a grind: copy, Alt-Tab, paste, switch back, repeat.
What came out of it
SafeRelay. Highlight text anywhere, press one hotkey, and it relays into the right place — AI tool, editor, ticket, terminal — in milliseconds, with the sensitive parts stripped out locally on the way. Fast enough to stay in your flow. Safe because the redaction never leaves your machine. Accurate because it's built on purpose-made detection patterns, not guesswork.
It shipped across the browser, VS Code, the CLI, Docker, an SDK, and a signed Windows desktop agent — on a hardened CI/CD pipeline.
What LogicGrid builds
Practical infrastructure and workflow tools for people and organizations that need speed, accessibility, and security at the same time — spanning secure AI workflows, local-first data protection, WhatsApp-first services, and developer tooling.
The throughline is simple: build security in, ship it for real, and let nothing leave the machine that shouldn't.