From Discovery Call to a Tiered Proposal — Systematically
An automated pipeline that turns messy call transcripts into a clear, three-tier solution map.
The problem
Why this needed to exist
The slowest, most inconsistent part of selling a technical service is the gap between the discovery call and the proposal. You finish a great conversation, then sit down to a pile of notes and try to turn it into something a busy owner can actually decide on. Do it by hand every time and the quality drifts, the turnaround drags, and good options get forgotten.
Small-business owners also don't want a single take-it-or-leave-it quote. They want to see options — what they could do themselves for free, what's worth paying for, and what a custom build would take.
The approach
How I built it
I built an assessment pipeline that ingests Otter or Fathom transcripts, aggregates multiple employees' calls for the same company, and surfaces where their stories contradict each other (which is usually where the real problems hide). From there it extracts the underlying problems, ideates solutions, and maps each one onto three tiers: free/DIY tools, productized services, and custom builds.
The output isn't a raw AI dump — it's a structured Solution Map plus a proposal seed, a 5C-style audit, and an outline for the walkthrough video. The automation does the heavy lifting; I keep the judgment call on what to actually recommend.
- Transcript ingestion (Otter / Fathom)
- Multi-call aggregation
- Tiered solution mapping
- Proposal + audit generation
The outcome
What it actually does
Every engagement now runs through the same repeatable path: transcript in, structured assessment out. It compresses the call-to-proposal time, keeps the quality consistent from one client to the next, and catches contradictions across a team that a human skimming notes would miss.
It's been run against real engagements across several verticals — roofing, insurance, real estate, therapy, and cleaning — producing solution maps tied to a clear three-option pricing structure.
- Transcript → Solution Map → proposal seed → audit, as one workflow
- Aggregates multiple employees per company and flags contradictions
- Maps every problem to a free / productized / custom tier
- Used on live engagements across multiple industries
What I learned
The curve I already climbed
The hard part wasn't getting an AI to summarize a call — it was designing the tiering so recommendations sort cleanly by technical complexity versus business value, and building automation that accelerates the work without quietly removing the human judgment a good proposal depends on.
Having built and run this on real clients means I already know where transcript automation gets things wrong and where it shines. A business adopting this cold would spend weeks learning those boundaries; I can apply a process that already works.
You get a fast, consistent assessment-to-proposal process that's already been pressure-tested on real engagements — not an experiment run on your dime.
Want this kind of thing built for your business?
Book a quick call and I'll show you what would actually move the needle for you.