NextFlow Builder Book audit →
// detailed scope

Packages, deliverables, and the lines we draw.

Below: what each mission includes, how long it runs, what falls outside the scope, and what stays under your control. All prices in USD. Software subscriptions (n8n cloud, Zapier, LLM credits, Workato, etc.) are billed to your accounts, not mine.

Send a brief→ Compare missions
// mission compare

Three missions, side by side.

Audit lands cheap and fast — most teams discover 2–3 quick wins worth a sprint. Sprint is the workhorse. Setup is for teams ready to wire an AI agent or MCP integration into production.

M-01 Audit 5 days

AI Automation Audit

$499

For teams who suspect 30–50% of their ops work is automatable but don't know where to start.

  • 2 × 30-min discovery calls with the ops/RevOps lead
  • Read-only access review of your existing stack
  • Time-leak ranking — which workflows cost the most hours/month
  • 30-day automation roadmap, ranked by ROI and risk
  • 1 quick-win blueprint, ready to hand to any engineer
  • 20-min Loom walkthrough on delivery
  • $499 credited toward a Sprint or Setup if booked within 30 days
Book audit→
most-shipped
M-02 Sprint 10–14 days

Workflow Sprint

$1,750

Build 2–3 production workflows on n8n or Zapier, with one AI-powered step. Most teams start here after the audit.

  • 2–3 production-ready workflows (n8n self-hosted or Zapier)
  • 1 AI step — classification, summarization, routing, drafting, or extraction
  • Real integrations with 3–5 tools in your stack
  • Error handling, retries, and dead-letter pattern
  • Telemetry hooks (Slack / email / dashboard alert on failure)
  • Documentation + 20-min Loom walkthrough
  • 14-day post-ship support for fixes and tweaks
  • 50/50 billing — half on signing, half on ship · Wise / AirTM / wire / PayPal
Start sprint→
M-03 Setup 3–4 weeks

AI Agent / MCP Setup

$3,500

One AI agent or MCP integration set up on your stack — and shipped to production with proper observability.

  • Either: an AI agent with multi-tool orchestration
  • Or: an MCP integration — official MCP servers (Zoho, GitHub, etc.), MCP-native platforms (Workato), or native LLM tool-calling (OpenAI / Anthropic function calls)
  • Prompt design + tool definitions + evals on representative sample inputs
  • Error handling, retries, fallback paths, rate-limit guards
  • Observability — request logs, token costs, latency, success rate
  • Connect 2–3 tools from your stack (HubSpot / Linear / Notion / Slack / Intercom / Stripe / etc.)
  • 30-day post-ship support · one iteration round after initial production usage
  • 50/50 billing — half on signing, half on ship · Wise / AirTM / wire / PayPal
  • Custom MCP server build · add-on starting at $1,500 (when an existing server doesn't exist)
Scope setup→
// M-04 · optional retainer

Automation Care.

An ongoing automation engineering channel for teams that want a senior pair of hands without the headcount math. Best after at least one Sprint or Setup so the stack context is already loaded.

// scope

What's included monthly

  • 6–8 hours per month of senior automation + AI work
  • Maintenance on existing workflows and agents
  • New builds within the monthly hour budget
  • Failure monitoring + first-response within 1 business day
  • Monthly written summary of work completed
  • Priority on incoming requests over non-retainer clients
// terms
$1,200/mo
  • 3-month minimum to start
  • Monthly cancel after the minimum (30 days notice)
  • Pause up to 60 days without penalty
  • Hours don't bank — buys ongoing capacity, not credit
  • Hour overruns billed at $125/hr, pre-approved
  • Monthly invoice · Wise / AirTM / wire / PayPal
Talk about care→
// boundaries

What's in scope, what isn't, what you provide.

Saving everyone the awkward mid-project conversation. If something here isn't workable, mention it before signing.

// included in delivery

  • Discovery and scope confirmation before any build
  • All configuration, code, and prompts for in-scope workflows
  • Test runs with sample payloads before production deploy
  • Handoff documentation for everything shipped
  • Post-ship support per the package's support window
  • Source code transferred to your repo on completion

// billed to your accounts

  • n8n cloud / self-host infrastructure costs
  • Zapier paid plan if needed for the workflow tier
  • Workato licenses (if using as MCP-native platform)
  • LLM API usage (OpenAI, Anthropic, Gemini tokens)
  • HubSpot, Salesforce, Linear, Notion, Slack paid tiers
  • Third-party API credits (Stripe, Twilio, SendGrid, etc.)
  • Domain, hosting, observability tooling (Sentry, Posthog, etc.)

// you provide

  • Access (or scoped credentials) to the tools we'll automate
  • A point-of-contact who can confirm requirements and sign off
  • Description of the workflows in plain language up front
  • Sample data — at least 5 representative inputs per workflow
  • Decision on any policy/compliance edges before build
  • Approval on production deploy

// not in scope

  • Full-stack SaaS product development
  • Mobile or desktop application builds
  • Marketing-site design (we have one, but it's not the service)
  • Paid-ad campaign management or social-media operation
  • 24/7 on-call rotation (SLAs discussed for retainer clients)
  • HR / sales / financial process consulting beyond the automation surface

// IP & ownership

  • You own all work product — code, prompts, recipes, docs
  • No proprietary lock-in; everything ships to your repo and tools
  • NDA on request — yours or a clean mutual NDA template
  • Aggregate, anonymized learnings inform future work (no client data)
  • Open-source dependencies keep their original licenses

// technical baseline

  • Built on documented, supported features of each platform
  • No reliance on undocumented or private APIs
  • n8n: trigger nodes, action nodes, webhook, HTTP request, custom code
  • Zapier: triggers, actions, paths, filters, webhooks, code steps
  • MCP: official servers (Zoho, GitHub, etc.), MCP-native platforms (Workato), or custom builds via @modelcontextprotocol/sdk
  • LLM: function/tool calling, structured outputs, eval-driven prompts
// next step

Ready to brief a mission?

If you've made it this far, you've read more about my scope than 95% of inbound. The shortest path to a yes-or-no is a 5-minute brief.

Send a brief→ Back to homepage