Pricing

Fixed fee. Clear scope. Built to ship.

Three common ways in, plus enablement for teams that want to own the next connector themselves. We scope the work, you approve it, then we build against a written definition of done.

Scope an engagement → Compare options
◎01 / Audit

$8k

Two-week diagnostic for teams deciding where MCP belongs in their stack.

  • Data and tool inventory
  • MCP architecture sketch
  • Security and auth notes
  • Prioritized 90-day roadmap
Start with audit →
Most teams pick this
▥02 / Build

from $35k

Four to eight weeks to ship production MCP servers on your infrastructure.

  • 2-4 production connectors
  • Scoped auth and audit logs
  • Eval harness and CI
  • 30-day post-ship support
Scope a build →
↗03 / Retain

$12k/mo

Ongoing MCP engineering for teams with a roadmap longer than one launch.

  • Fractional senior engineer
  • New connector delivery
  • Quarterly strategy reviews
  • Slack or Teams channel
Talk retainer →
◌04 / Enable

custom

Workshops and paired implementation for teams that want to own MCP internally.

  • Architecture workshop
  • Hands-on implementation lab
  • Repo templates and patterns
  • Internal launch support
Plan enablement →
How to choose

Pick the smallest engagement that removes the most uncertainty.

Good MCP work should create leverage quickly without creating a new platform your team cannot own.

Engagement Best when Outcome
Audit You need the plan before the build. Architecture, roadmap, and clear scope.
Build You know the first systems agents need. Production connectors with auth, logs, evals, and docs.
Retain The roadmap spans multiple teams or quarters. Senior implementation capacity without a full-time hire.
Enable Your engineers want patterns they can repeat. Workshops, templates, and paired delivery.
FAQ

Do you work inside our infrastructure?

Yes. AWS, GCP, Azure, Vercel, on-prem, or a locked-down enterprise environment. The goal is for your team to own the deployed system.

FAQ

Which agents and clients?

Claude Desktop, Claude Code, Cursor, Gemini, OpenAI-compatible clients, and internal agent surfaces that speak MCP or can be adapted cleanly.

FAQ

Can we start smaller?

Yes. Start with the audit when the first connector is not obvious. Start with build when it is.

Ready?

Bring a messy stack. Leave with a plan.

Tell us what agents need to reach, what can go wrong, and who has to own it after launch.

Book a 30-min audit → See case studies