HermIIS

Engineering context disappears between planning and shipping.

Tasks live on a board, product intent lives in feature descriptions, architecture decisions live in documents, and the actual implementation lives in GitHub. Teams spend time rebuilding the relationships between those sources whenever priorities change or someone asks why a decision was made.

Make the relationships part of the workspace.

HermIIS treats project artifacts as connected product knowledge instead of unrelated pages.

• Plan work through customizable kanban boards, features, tasks, subtasks, roles, and templates.
• Capture architectural decisions with alternatives, version history, and supersession chains.
• Maintain versioned Markdown documents with diffs, anchors, cross-references, and stale-reference detection.
• Connect GitHub repositories so commits and pull requests remain linked to the work that motivated them.
• Explore tasks, features, decisions, documents, code activity, and dependencies through a knowledge graph.
• Measure delivery with burndown, velocity, cycle time, forecasting, reports, exports, and API access.

AI works from project evidence and shows its sources.

HermIIS can generate contextual task proposals from a feature and its connected codebase, answer natural-language questions across project knowledge, surface feature risks, and suggest when an architectural decision should be recorded.

Generated tasks, answers, reviews, and decision suggestions remain visible proposals. Citation chains and connected source records help the team inspect the reasoning before accepting work or changing direction.

A product model broad enough to stay connected.

I designed the domain model, multi-workspace permissions, kanban and feature workflows, decision and document versioning, knowledge graph, GitHub integration, AI retrieval and generation flows, analytics, reports, REST API, webhooks, PWA behavior, deployment, and public product experience.