Consulting on
agentic AI & automation.

Most of what Flow Local runs as a managed service can also be designed, built, and handed to an internal team. If you'd rather own the system than rent it, this is the other side of the business — audits, builds, and training for teams with the appetite to run it themselves.

No pricing listed on purpose: scope varies too much. A 15-minute call is the best way to tell whether this is the right shape of engagement for what you're trying to do.

Book a 15-minute call

Three shapes of engagement.

1. Audits

A proper look at how your current systems handle lead response, reviews, communication, acquisition, and reporting — where the leaks are, what's automatable, what isn't, and what order to fix things in. You get a written audit with specific recommendations and the rationale behind each one. Useful whether you intend to run the fixes yourself or hand them to a partner.

2. Builds

Designing and building a specific system — an agentic workflow, a lead-response pipeline, a review engine, a reporting stack, an internal tool. Scoped project, clear deliverable, handed over working with documentation. Most builds end with a training session so the team can run and extend what's been shipped.

3. Training

Bringing a team up to working competence on agentic AI and the automation stack that sits around it — what's real, what isn't, where the sharp edges are, and what to build first. Runs as a workshop or as an ongoing advisory pattern. Aimed at operators and team leads, not executives looking for a keynote.

Platforms & stack.

Not every project uses all of this — most use a handful. The point is that tool choice is driven by the problem, not by being locked into a single vendor.

Models & agent frameworks. Claude (Anthropic), OpenAI, open-source models where they fit. Agent frameworks and MCP servers for structured multi-step work. Claude Code and Cursor for build and iteration.

Automation & orchestration. n8n, Make, Zapier, Pipedream. Choice depends on scale, who'll maintain it, and what needs to run on-prem vs. hosted.

CRM & communication. HubSpot, Pipedrive, Salesforce, GoHighLevel (GHL), Attio. Twilio, Vonage, and messaging APIs where the CRM's native comms aren't enough. Calendly, Cal.com, and custom booking where scheduling is load-bearing.

Data & reporting. Supabase, Postgres, BigQuery. Looker Studio, Metabase, Mode. Supermetrics and Funnel.io for ad-platform pipes. PostHog and Mixpanel for product analytics.

Frontend & internal tools. Next.js on Vercel, Retool and Tooljet for internal UIs, Framer and Webflow for marketing surfaces where appropriate.

Telephony & voice. Vapi, Retell, ElevenLabs, and the major telephony providers for voice-agent work when it's actually the right answer — which is rarer than most vendors suggest.

Best way forward is a 15-minute call.

Bring the problem you're trying to solve. If it's a fit, we'll scope it properly. If it isn't, I'll say so and point you somewhere useful.