What You Get
- A constitution for your project, encoded as steering rules your agents actually load
- Spec Kit installed and running on your repository — the full command loop, not a demo
- Engineering and product trained together, on your codebase, in the same room
- A golden test set and an offline eval harness your team can re-run every Bolt
- Review gates wired into CI, with the human/machine boundary written down
- One real Unit of Work shipped to production under the new process
- A written operating manual: ceremonies, artifacts, approval gates, escalation path
Packages
Agentic Bootcamp
from $6,000
One week. Engineers and product managers fluent in the loop and the vocabulary.
- AI-DLC phases, Bolts, Units of Work
- Full Spec Kit command loop, hands-on
- Constitution drafted for your project
First Bolt
from $18,000
Bootcamp plus a real Unit of Work shipped to production under the new process.
- Spec Kit installed on your repo
- Eval harness + review gates in CI
- One Unit of Work shipped, not simulated
Full Enablement
from $45,000 · coaching from $4,500/mo
The complete install-and-hand-over programme across a multi-team organisation.
- Assess → Install → First Bolt → Hand over
- MCP integration across your systems
- Operating manual + optional embedded coaching
In Scope
Assess
Week 0- Delivery baseline: cycle time, test coverage, version-control maturity
- Audit of where AI has already entered the team unmanaged
- Readiness call: enablement now, or fundamentals repair first
- Track selection — which squad and which Unit of Work goes first
Install
Weeks 1–2- Constitution written and encoded as steering rules (AGENTS.md and agent-native equivalents)
- Spec Kit initialised against your repository and your coding agent
- MCP servers wired to your repo, database, and cloud with scoped, audited permissions
- Golden test set, offline eval harness, and review gates defined in CI
Run the first Bolt
Weeks 3–4- Mob Elaboration on a real Intent, with product and engineering in the room
- Decomposition into Units of Work with dependencies and parallelisation boundaries
- Spec → clarify → plan → tasks → analyze → implement, on your codebase
- Ship to production behind your existing review and release process
Hand over
Weeks 5–8- Your team runs Bolts unaided while we observe and audit
- Ceremony redesign: what replaces grooming, planning, and retro
- Written operating manual — artifacts, approval gates, escalation path
- Exit review, with optional embedded coaching from month two
How It Works
01
Assess — we baseline your delivery fundamentals and tell you honestly whether agentic throughput is safe on them yet.
02
Install and run — constitution, Spec Kit, evals, and review gates go in; then one real Unit of Work ships under the new process with us alongside your team.
03
Hand over — your team runs Bolts without us, we audit and coach, and you keep every artifact in your own repository.
What “Done” Means
Your team plans, elaborates, and ships a Bolt end to end without a Codenovai engineer in the room.
Why Codenovai
We do not teach this from slides. Every session runs on your repository, and the programme is not finished until one real Unit of Work from your roadmap has shipped to production under the new process — with your product manager running the elaboration, not ours.
Fixed price
Shipped in weeks
Code + IP in your repo, day one
No lock-in
FAQ
- What is AI-DLC?
- AI-DLC (AI-Driven Development Lifecycle) is an open-source, AI-native methodology from AWS Labs. The AI drafts a detailed plan, asks the questions it cannot answer, and defers every critical decision to a human before acting. Work moves through three phases — Inception (what and why), Construction (how), and Operations (deploy and run) — each elaborated by the whole team together rather than passed between silos.
- What is a Bolt, and why does it replace the sprint?
- A Bolt is AI-DLC's smallest iteration, measured in hours or days rather than two weeks. The sprint exists because elaboration, building, and review each needed their own calendar slot. When the AI drafts the plan live and implements against an approved spec, that batching stops being necessary — so the iteration shrinks to match real cycle time. Epics become Units of Work for the same reason.
- How does Spec Kit fit alongside AI-DLC?
- AI-DLC gives you the methodology and the team rituals; GitHub's Spec Kit gives you the mechanism inside the Construction phase — concrete commands (constitution, specify, clarify, plan, tasks, analyze, implement) that turn a specification into an executable artifact. We teach them together because each covers the other's gap, and adopting one alone is the most common failure we are called in to fix.
- Do you train product managers as well as engineers?
- Yes, and we will not run the programme engineering-only. Mob Elaboration needs the person who can decide what the product should do, answering the AI's clarifying questions live. Product managers learn to write Intents instead of tickets, decompose them into Units of Work, write machine-verifiable acceptance criteria, and read eval output in place of story-point burndown.
- Do we have to switch to AWS, Kiro, or GitHub?
- No. AI-DLC is explicitly agnostic to IDE, agent, and model, and Spec Kit supports 30-plus coding agents. We install both against whatever you already run — your cloud, your repository host, your model provider. Model independence is a design requirement in every Codenovai engagement.
- Does this work on a legacy codebase?
- Yes, and brownfield is the more common engagement. Spec Kit's converge command assesses an existing codebase against your artifacts and reports what is genuinely outstanding, which gives you a real starting position instead of a rewrite fantasy. On legacy work we spend more of the Assess week on test coverage, because that determines whether agentic delivery is safe there yet.
- We already run Scrum. Do we throw it out?
- No, and you should not start by trying. Teams keep the intent behind their ceremonies and change the mechanism: grooming becomes Mob Elaboration, planning collapses into the Bolt itself, and retro becomes a Bolt review whose output is a correction to the constitution or the spec. We migrate one squad first, measure it, then decide about the rest.
- Who owns the artifacts afterwards?
- You do, from day one. The constitution, specs, steering files, eval harness, and every generated artifact live in your repository. AI-DLC is open source from AWS Labs and Spec Kit is MIT-licensed from GitHub, so there is nothing proprietary to Codenovai in the process you are left running. Independence is the deliverable.