AI is free.
Implementation
is the work.
Two clear tracks. We build it for you, or We teach you to build it yourself. The teach track costs a fraction of the build track because you do the work — we bring the blueprint, the failure modes, and the hours of debugging you skip. All prices starting from. Final scope agreed in writing after a free 30-minute discovery call.
Two ways to get there.
Discovery call · always free
Build With Me.
We take it end-to-end. You get a working sovereign AI mesh stood up in your stack within weeks. Premium price because you get premium speed and zero rework.
- · We write the code, run the deploys, train your team
- · Production environment in 4–8 weeks
- · 30-day support window included
- · From £1,250 (architecture) to £7,500+ (full build)
Teach Me.
You do the work. We bring the architecture, the playbooks, and the debugging hours you skip. A fraction of the cost — fits founders, technical teams, and anyone who wants to own the build.
- · Workshops, mentored builds, office hours
- · You ship it. We keep you unstuck.
- · Weekly checkpoints, async support between
- · From £300/month (office hours) to £2,000 (mentored build)
Not sure which fits? Take the 30-minute discovery call. We'll figure it out together — at no cost.
Done for you.
Operator · Lee Akpareva, named
Design Your Mesh
- We review your business, stack, and constraints
- We design the mesh that fits your size and goals
- Cost model + 90-day implementation roadmap
- Vendor lock-in audit on what you have today
- Final report + 60-min walkthrough with you
Build the Mesh
- We build your sovereign AI mesh end-to-end
- Cloudflare + Tailscale + Docker stack stood up
- Resident AI agent configured to your domain
- Identity-aware portal, observability, secrets vault
- Team training, complete runbook, on-call handover
- 30-day production support window included
Run the Mesh
- We keep your mesh up, current, and secure
- Monthly architecture review + optimisation
- Resident AI agent kept current as new providers ship
- 24/7 incident response with named operator
- Quarterly board-grade reporting
- Two scheduled feature additions per quarter
Done with you.
You ship it · we keep you unstuck
Two-Day Intensive
- The full architecture, end-to-end, with no skipped pieces
- Live walkthrough of the reference network
- You leave with a draft architecture for your business
- Up to 6 attendees from your team
- Recording + slide pack to keep
- One-week async follow-up window
We Guide. You Build.
- You implement, we unblock
- Weekly 90-minute working session, async between
- Architecture decisions reviewed before commit
- Code review on the parts that matter most
- You ship it. You own the IP. You learn it.
- End-of-engagement runbook handover
Ongoing Access
- For teams who have built it and want a sounding board
- Async questions answered within one business day
- One scheduled 60-minute call per month
- Architecture review on request
- Cancel anytime after the third month
All Track B engagements include access to the same blueprint we use to operate the reference network.
Honest, open to negotiation.
No SDR pricing games
All numbers above are starting from. They reflect the typical entry point for the work described. Genuine scope is agreed after the free discovery call and confirmed in writing before any commitment.
Early-stage founders, non-profits, regulated sectors with public-good outcomes, longer engagements, equity-for-work — all worth a conversation. Bring your context, we will meet you where you are.
A small, direct engagement model.
Time-to-start · < 2 weeks
One conversation, then a plan.
Every engagement starts with a free 30-minute discovery call. We look at your stack, what is hurting, what is locking you in, and which track fits — Build With Me or Teach Me. If we are not the right fit, we will say so and point you at someone who is.
Working software in week two.
For Track A, no multi-month design phase. By the end of week two, you have a working environment we can demo. For Track B, by the end of week two you have shipped your first node — together.
You are never alone with it.
Every engagement includes a follow-up window. Most clients move into office hours or a managed retainer afterwards — same operator, same context, same chat thread.
The architecture grows with you.
Add nodes, add agents, add tools — the architecture is designed to be extended, never rewritten. Capabilities ship as containers and skills, not as new platforms.
Common questions.
How do I choose between Build With Me and Teach Me?
If you have technical capacity in-house and want to own the implementation, take Track B (Teach Me). You will save 60–80% on the cost and your team will deeply understand the system. If you want it shipped fast, with no learning curve on your side, take Track A (Build With Me). Most clients start with a workshop (B1) and decide from there.
Are the prices negotiable?
Yes. The numbers on this page are starting points for the typical scope. Genuine scope and price are agreed after the free discovery call and confirmed in writing. Founders, non-profits, public-good projects, longer engagements, and equity-for-work are all welcome to start a conversation.
Will the network be hosted with you, or with us?
Either. Track A defaults to deploying into your cloud accounts and on your hardware — you own the keys, the data, and the bill. We can also host it for you under the Managed Operations tier. Track B always deploys into your environment.
Do we need a DevOps team to run this?
No. The whole point of the architecture is that one operator can run it. Most clients keep me on retainer for the first 6–12 months, then take it in-house with a single platform engineer.
How is this different from buying Vercel, AWS Bedrock, or Anthropic's API?
Those are excellent products. NAVADA Edge is the assembly that sits above them — the orchestration, the agents, the data plane, the observability, and the day-to-day operations. We use Bedrock and the Anthropic API inside the network. We just do not depend on any one of them.
What does the run-rate look like for a client environment?
The reference network operates at roughly £70 per month, all-in. A client environment with redundancy, higher SLAs, and dedicated GPU capacity typically lands between £400 and £1,250 per month depending on workload. We model it explicitly during Discovery.
What if our compliance team blocks public cloud?
The Edge Server reference build is designed exactly for this. Workstation-class hardware in your building, on your network, mesh-joined to whichever cloud regions you do permit.