Bootcamp keeps your documentation, tasks, rules and AI agents in one governed workspace, from software builds to ERP operations, so work moves forward and nothing gets lost.
Specs, decisions and guides sit next to the work, not in a forgotten wiki.
about 500 governed rules and quality gates across 50 rule sets keep every agent and every human on the same track.
Build on the ERP's APIs instead of customising it. Agents work with the user's permissions, a human reviews every risky write-back, and everything leaves an audit trail. The agent prepares; the ERP posts.
Connect Cursor, Claude Code or your own assistant through Bootcamp's MCP server, under your team's rules.
Development, migration, ERP analysis (Fit-Gap), IT management and CAD/CAM management.
No login, about five minutes. Spoiler: step 6 is where your own bot gets a seat at the table.
Start the demo →Without structure, decisions disappear, scope grows unnoticed and nobody can say what the agent actually did.
Specs and decisions live in a wiki that nobody updates. By the next sprint they don't match reality — and the agent has no single source of truth.
Work is agreed in Slack or email, then lost. There is no board with owners, priorities, or a clear path from idea to done.
Agents ship fast, but without shared rules nobody can tell whether the output is safe to merge, document, or show to the business.
Documentation, governed tasks, roadmap and your own bots — connected under the same rules, for software builds and ERP operations alike.
Feature specs with acceptance criteria, decision records (ADRs), guides and a What's new log, browsable in the portal by anyone. Updated as part of every change.
Task Center board (Backlog → Elaborating → Spec review → Accepted → In progress → Review → Closed), each card with priority, owner and linked spec. Nothing gets built before a human accepts it; the agent builds, tests and documents, and the merge stays with you.
Roadmap that reflects what is actually being built, rules pushed to every project, compliance checks.
Bootcamp's MCP server: connect Cursor, Claude Code, VS Code or your own assistant; it signs in as you (or uses a personal key you can switch off), reads docs, updates task cards, comments and writes specs and decision records under the same rules.
Stripe-based checkout and webhook handling for subscription billing.
Last updated 2 hours ago · Spec phase 2
Acceptance criteriaBootcamp's documentation isn't written once and forgotten. Every feature spec, decision, and roadmap is a living document — updated as part of every change, visible to your whole team in the portal.
Connect any MCP-capable assistant to your Bootcamp project. It sees the same docs and tasks as your team, and it plays by the same rules.
✓ Fit-Gap analysis
✓ Make / buy / customise decision (ADR)
✓ Side-by-side API readiness
⌛ Agent write-back: waiting for human review
Build on the ERP's APIs instead of customising it. Agents work with the user's permissions, a human reviews every risky write-back, and everything leaves an audit trail. The agent prepares; the ERP posts.
Any team member can interact with their project through Bootcamp's chat — ask questions, request updates, kick off tasks, or review what's happening right now. No command line, no code.
See how structured AI development shifts outcomes across your whole team.
| Area | Without Bootcamp | With Bootcamp |
|---|---|---|
| Management visibility | Updates only when asked — and often incomplete | Live delivery status, lifecycle stage, and cost tracking at all times |
| Scope control | Features grow mid-sprint without anyone noticing | Every feature requires a spec with acceptance criteria before work starts |
| AI development quality | Fast but unpredictable — needs heavy human review | Agents follow the same rules and quality gates as people |
| Your bots | Copy-paste between chat and the backlog | Connected through MCP, working on the same tasks and docs |
| Documentation | Written once, outdated within weeks, trusted by nobody | Updated with every change — always reflects the current state |
| Decision traceability | Verbal, forgotten after the meeting — no record | Every decision is documented, searchable, and audit-ready |
| Delivery confidence | Hope-driven deadlines with no structured milestones | Incremental delivery with go/no-go criteria and team sign-off |
The Bootcamp portal gives every role their own view — chat, live docs, roadmap, cost tracking, and code preview — all from a browser. No IDE needed.

Ask questions, request summaries, or kick off tasks in plain language — no technical knowledge required.

Browse feature specs, decisions, and roadmaps directly in the portal — always up to date with the codebase.

Management and product see exactly which lifecycle stage the project is in, what's done, and what's next — in real time.

About 500 governed rules and quality gates across 50 rule setscover quality, testing and process — applied automatically to every project and every AI agent.

Developers and AI agents work from the same specs and standards — in Cursor or any IDE — producing predictable output.

Track human and AI effort per project — shown as hours, not tokens. A language your team actually understands.
Pick the kind of work you're running. Structure, documents and rules are ready on day one.
Web apps, APIs, desktop, multi-stack repos.
Legacy to modern stacks, incremental cut-overs.
Target stacks: Next.js · Full-stack (ASP.NET + React) · ASP.NET (Microsoft stack)
Fit-Gap, make/buy/customise, API readiness, agentic add-ons next to the ERP.
Discovery, runbooks, compliance gap analysis.
AutoCAD, Inventor, TopSolid and Blender workspaces with Check out / Check in on shared engineering files.
One plan. €25 per user per month, minimum 3 users. Start with a template, connect your own bots, keep every decision.