What was said
What this says
As at Fri 9 Oct 2026Tobias Renwick and Nell Ashcombe demoed a test brain on Zendesk, Salesforce and Xero data. Kenny Lockhart called it really useful and was happy with what he saw. A proposal is the next step, and Moira will want to see it.
What happened: Nell and Tobias showed Kenny Lockhart a test instance built on a subset of the customer's data, with unified search and natural-language questions. Moira Sinclair was not on this call.
Customer view: Kenny asked about freshness and whether staff could stay in their normal tools. He said, "That's actually really useful."
Commercial: Nell said a proposal would follow with scope, timeline and commercial terms, and quoted four weeks for implementation. Kenny asked about adding product usage data later, which points to more connectors.
Watch: Kenny said Moira will want to see this too, so the finance decision is still ahead. The refresh is nightly, and Nell promised the summary by end of that week.
Themes
As at Fri 9 Oct 2026Who was there
BG-Demo
Them
Follow-up
Next steps
As at Fri 9 Oct 2026- Nell Ashcombe to send Kenny a summary of what was built by the end of that week.
- Prepare a proposal with scope, timeline and commercial terms, and include Moira Sinclair.
Risks
As at Fri 9 Oct 2026- Moira Sinclair has not seen the demo and the finance decision is still open.
- The demo used a subset of data, so expectations at go-live need managing.
Opportunities
As at Fri 9 Oct 2026- Adding a product usage connector later would deepen the brain beyond the first connectors.
- A Starter proposal can follow while Kenny is positive.
Transcript
- Nell AshcombeHi Kenny. Tobias is here as well, he's our CTO and built the platform. Thanks for getting your data to us. Ready to see what we did with it?
- Kenny LockhartYeah, really keen to see it.
- Tobias RenwickSo what I've done is build a test instance of the brain on your Zendesk, Salesforce and Xero data. It's running in isolation, this is just for the demo, your live data is safe. Let me share the screen.
- Nell AshcombeSo Kenny, what you're looking at is a unified search across your customer support data. All your recent tickets, the customers who raised them, and whether they're paying. Let me show you what I mean. If a support person gets a ticket from a customer: let's call them Acme: and they search for Acme in the brain, they'd see this. Ticket history, account status, payment status, who they've spoken to before. All in one place.
- Kenny LockhartMm. And that's pulling from all three systems?
- Tobias RenwickTo be precise, yes. We've indexed your Zendesk tickets, your Salesforce accounts and contacts, and your Xero invoices. When you search, the brain queries all three in parallel and presents a unified view.
- Nell AshcombeAnd you can ask it questions in natural language. So if your team wanted to know 'show me all customers with unpaid invoices who've reported a ticket in the last week', the brain would search all your data and answer that.
- Kenny LockhartRight, and how fresh is the data?
- Tobias RenwickWe refresh every night. So a ticket entered in Zendesk this afternoon will be searchable in the brain by tomorrow morning.
- Nell AshcombeLet me show you another example. One of the common questions for support is, why is this customer angry? Is it a known issue, have they complained before, are they an important account? Watch this. I'll search for a specific ticket and ask the brain to summarize the customer's history.
- Tobias RenwickAnd the brain returns a summary: this customer has reported three issues, all related to integrations, all in the last month. They're on a Growth plan, they've been paying on time. So this is a technical issue, not a commercial one.
- Kenny LockhartThat's actually really useful.
- Nell AshcombeNow, this is a test build, we've used a subset of your data. But when you go live, it would have all your tickets, all your customers, all your history. And your team can also build saved searches. So if you want 'high-priority unpaid customers' to be a report your team checks every morning, you can do that. It runs every night and they see the answer first thing.
- Kenny LockhartAnd they'd be able to use this from where they are, not having to come to some other tool?
- Nell AshcombeExactly. They're in Zendesk, they open a new tab, they search the brain, they get the answer. It's browser-based, it's alongside their normal tools.
- Tobias RenwickOne more technical point, the honest answer is that your data is stored in your Snowflake account, which you already own. We don't replicate it to our servers. We run the brain inside your warehouse. That means you keep control, and it means you can add other data sources, HR systems, product usage data, anything you put in Snowflake, and the brain can search across that too.
- Kenny LockhartSo if we added, say, our product usage data later, the brain could connect it?
- Tobias RenwickYeah. You'd add that connector and the brain would index it alongside your support and account data.
- Kenny LockhartOK. That's good. So what happens next?
- Nell AshcombeI'll send you a summary of what we've built. Then we'd move to a proposal: we'd lay out the scope, the implementation timeline, and the commercial terms. You'd have a chance to ask questions about anything, and if you're happy, we'd start the real build.
- Kenny LockhartHow long is the implementation?
- Nell AshcombeFour weeks, with some effort from your side. We'd need someone: could be you: to validate the data, make sure we haven't missed anything, and then we'd do a week of training with your support team.
- Kenny LockhartAnd after that it's all live and the team can use it?
- Nell AshcombeYeah. We'll check in a couple of times in that first month to make sure everyone's confident. Then after that it's part of your normal toolkit.
- Kenny LockhartOK, cool. I think Moira will want to see this too, but I'm pretty happy with what I'm seeing. When can you send that summary?
- Nell AshcombeI'll get it to you by end of this week.
- Kenny LockhartGreat. Cheers, you two.