Notes from the workshop.

Auptim today announced the private beta of Shmidge, an AI project system for small software teams that are tired of keeping the board alive by hand. Shmidge brings issues, epics, sprints, workflows, chat, GitHub activity, proposals, reporting, Vega Assistant, and coding-agent-ready project context into one place.

Small software teams often hit the same wall: requirements scatter across chat and calls, blockers hide in pull requests, and status depends on whoever remembered to update the ticket. Shmidge is built for that gap between planning and implementation, where the real work happens faster than the tool can describe it.

Inside Shmidge, Vega Assistant can turn rough notes into tickets, edit stale issues, search the backlog, judge priority, move work in bulk, create sprints, summarize chat, and find the old conversation that explains why a decision was made. Since Shmidge connects through GitHub, Vega can also inspect indexed repositories, search files, read code snippets, diagnose where a problem lives, and help draft tickets with real code context.

Shmidge Companion carries that context into coding-agent work. Claude Code and Codex can pull board context, create or update tickets, create epics and sprints, link branches and commits, and leave tracked work sessions behind the changes they touch. The result is a project system the agents can understand while the humans keep shipping.

For distributed teams, Shmidge also keeps language from becoming another handoff. Project chat, private assistant chat, tickets, and comments can translate into each teammate's preferred language, with optional Chrome AI translation for supported chat sessions and server translation for broader coverage.

The private beta is for small software teams shipping real product work and feeling the drag between planning, implementation, and reporting. Auptim is reading every application personally and onboarding the first cohort hands-on.

Apply at shmidge.ai/register.