AltSql
A native hybrid of key-value and SQL, from the sensor to the gateway. Both ends share the same records, byte for byte.
01/Studio
We design and build databases, proxies, queues and backend systems, for clients and in the open.
02/Work
Our own projects, all open source under Apache 2.0. Each one has its own site, and its code and releases are on GitHub.
A native hybrid of key-value and SQL, from the sensor to the gateway. Both ends share the same records, byte for byte.
SQL, key-value, graph and vector in one native database, kept in a single SQLite file.
A job queue in one file, with retries, delayed jobs and dead letters built in.
Answers kept ready as data arrives. Counts, percentiles and summaries stay current in a SQLite file.
The operating system for your web traffic: a small Go kernel with sandboxed WebAssembly plugins on top.
The VPN, rebuilt as an operating system. Each program gets its own network path, policy and record.
Ready machines for coding agents. Describe a project’s setup once, and preconfig writes each agent platform’s setup file and tests it on a clean machine.
Free tools for the key-value stores you run: Redis, Valkey, DynamoDB and etcd. Nothing you paste leaves your browser.
01
A working first version you can put in front of users or investors. We find the smallest build that proves the idea, and build that first.
02
Services and command-line tools, mostly in Go. Tests and release builds start with the first commit.
03
Embedded databases, key-value stores, job queues and precomputed views, sized for anything from a sensor to one honest server.
04
Reverse proxies, tunnels, traffic rules and plugin systems, including sandboxed WebAssembly plugins you can write in any language.
05
We join your team for a feature or a migration that has stalled, and work in your repo by your conventions.
06
Release pipelines, binaries for each platform, changelogs and docs, so a project keeps shipping without anyone babysitting it.
04/Process
The same four steps every time. You see working code early, and you see it often.
A call, then a short written brief that says what it must do and how we’ll know it works.
A small working build early on, so decisions get made on something you can run.
Short cycles, each ending in a version that runs. You see the repository the whole time.
Code, docs, binaries and an automated release pipeline. You own all of it, and we stay reachable.
Tell us what it has to do and what it runs on. We’ll reply with questions first, then a plan and a rough size.