Sales and business systems
A messaging engine that turns customer conversations into what reps should say next
A knowledge layer on a company's products, services and competitors, fed with call transcripts, customer emails and support tickets. Out the other side come value propositions, opportunity ideas and customer goals, written as clear messaging for every rep before their next call.
Most sales enablement is a slide deck nobody opens. The information reps actually need is scattered across call recordings, email threads and support tickets, in the customer’s own words. I built the thing that reads all of it.
The knowledge layer
First, the system had to know what the business sells. I built a structured knowledge base covering its products and services, what each one is for, how it is priced and packaged, and where each competitor is stronger or weaker. That is the reference every later step checks against, so the output talks about real capabilities rather than plausible-sounding ones.
What it reads
Three sources, all of which already existed and none of which anyone had time to read end to end:
- Call transcripts. Every discovery call, demo and business review.
- Customer communications. Email threads with each account.
- Support tickets. What customers complain about, ask for and struggle with after they buy.
The engine reads each source against the knowledge layer and pulls out the same structure every time: what the customer says they are trying to achieve, where they are stuck, which competitors they mention, what they think the business does well, and what they wish it did.
What it produces
For every account, three things a rep can use in the next call:
- Value propositions written in the customer’s own language, tied to a capability the business actually has.
- Opportunity development ideas with the evidence behind each one. Not “they might want X”, but “on the call in March they said they had a problem with Y, and ticket 4471 shows it is still open”.
- Customer goals, stated plainly, so the next conversation starts from what they want rather than what the rep wants to sell.
All of it lands as short messaging per persona, so the same insight reads differently for the CFO and the engineering lead, and a weekly digest for the sales lead showing which accounts have new signals.
What it changes
Reps stop opening calls with a pitch and start opening them with the customer’s own goals. That is the whole shift. The engine does not make anyone a better salesperson. It makes sure every rep walks in knowing what a good salesperson would already have known.
What broke
Transcripts are messy: two speakers labelled the same, product names transcribed wrong, and the occasional 40-minute call about the weather. The fix was a first pass that only cleans and labels, then a second pass that extracts, with the extraction told to quote the source line for every claim. If it cannot quote it, it does not get to say it.
The principle
The best sales content is not written by marketing. It is already sitting in your customers’ sentences. The job is to find it, structure it and hand it back to the people who talk to them.