Work

Real systems, running in production.

A selection of AI systems built and deployed by ResonateOps: two from client engagements and two from our own operation, because we run the same systems we sell. Client cases are anonymized by default; every claim here describes a system that is live today.

AI Sales Ops · Our own operation

Every sales call becomes a proposal, automatically.

Context

Writing a proposal after each discovery call took hours of replaying notes, and deals cooled while the document sat half-done.

What was built

A pipeline that takes the recorded call, extracts scope, pain points and pricing signals, and drafts a structured proposal filed against the right deal in the CRM. A human reviews and sends; nothing goes out on its own.

Architecture

Discovery callTranscriptAI extractionProposal draft (PDF)CRM deal

Stack

FathomPythonRailwayClaudeClickUp

Results

proposal drafts in minutes, not hours
a same-day proposal on every deal
human review before anything is sent

AI Outbound Engine · Our own client acquisition

The outreach system we sell, running our own pipeline.

Context

We needed a predictable flow of B2B conversations without buying ads or waiting on referrals — the same problem our clients bring us.

What was built

Dedicated sending infrastructure on secondary domains with warmup and authentication, verified prospect lists, and reply-focused sequences with deliverability monitored mailbox by mailbox. The same system we install as the Lead Acquisition Engine.

Architecture

ICP listVerificationWarmed inboxesSequencesRepliesBooked calls

Stack

InstantlyMillionVerifierPythonClaude

Results

thousands of verified prospects contacted
deliverability held healthy
replies land as booked calls

AI Analytics Platform · US ecommerce (under NDA)

A dashboard the CEO can finally trust.

Context

A US ecommerce brand whose command-center dashboard never matched the CEO's own weekly scorecard, so nobody trusted the numbers.

What was built

The root cause wasn't code: the scorecard blended two attribution sources into a single figure. I wrote a data contract, one clear definition per metric, and reconciled every number back to the source of truth, platform-attributed and consistent.

Architecture

Connected sourcesCollectorsData contract + calcDashboard = Scorecard

Stack

PythonFlaskSQLiteShopify / Meta / Google APIs

Results

numbers that reconcile
a definition behind each metric
a dashboard leadership trusts

AI Sales Platform · European sales consultancy

A sales methodology, turned into software.

Context

A sales consultancy whose value lived in the founder's methodology, trapped in decks and workshops, with nothing her clients could actually use between sessions.

What was built

I turned the methodology into a working product: a guided diagnostic that walks a client from symptoms to root causes, client workspaces, and an AI layer that helps structure and draft the thinking. Demo-first, shipped weekly; a one-off build grew into an ongoing retainer.

Architecture

Expert methodologyGuided diagnosticClient workspacesAI drafting layer

Stack

ReactTypeScriptSupabaseVercelAnthropic SDK

Results

a product clients log into
from one-off project to monthly retainer
visible progress every week

Client cases are anonymized under the reference clause of our service agreements. Details and references available on a call.

Want a system like these in your operation?

In 15 minutes I'll tell you straight which of these patterns fits your operation and what return to expect.