Dickey Singh Wants Customer Success to Run on Autopilot

Every customer success leader knows the trade that keeps them up at night. The account list grows. The team does not. So the top accounts get a human with a calendar and a quarterly review, and everyone else gets a newsletter and hope. Dickey Singh has spent most of his career staring at that trade, and with Cast.app he is arguing it was never a headcount problem in the first place.

Singh is the founder and CEO of Cast.app, a customer success platform for B2B software companies. The pitch is blunt enough to fit on a sticker: put net revenue retention on autopilot. Behind it is a specific bet about where AI actually earns its keep - not by making a salesperson type faster, but by showing up for the customer directly.

The problem he keeps returning to

If you trace Singh's résumé, the same thread runs through all of it: turning raw customer data into something a customer will actually act on. He held senior product and engineering roles at Silicon Valley firms including CustomerSat, MarketTools and ViVOtech, serving names like Apple, Google, Salesforce, SAP, MasterCard and Oracle. He founded the Encounters app, then co-founded and ran Pyze, a digital transformation analytics company out of Redwood City. He picked up an AI and business strategy program at MIT Sloan along the way.

Analytics companies are good at telling a business what its customers are doing. What they rarely do is close the loop - take that insight and put it in front of the customer in a way that changes behavior. Cast.app, founded in 2019, is Singh's attempt to close it.

Copilot AI for CSM teams and AI tools are table stakes. Autopilot AI is how you scale your business without scaling teams. Dickey Singh, founder and CEO of Cast.app

Copilot versus autopilot

This is the distinction Singh keeps hammering, and it is worth taking seriously because most of the market lands on the other side of it. The dominant story about AI in customer success is the copilot: a tool that drafts the follow-up email, summarizes the call, suggests the next step. Useful, and increasingly expected. But it still assumes a human in the loop for every account, which means it does nothing about the math that broke customer success in the first place.

Cast.app's product is built around what Singh calls Digital CSMs. They generate a personalized business review for an account - the health metrics, the adoption numbers, the value delivered - and then actually present it to the customer, answering questions grounded in the company's CRM, support and product data. It runs across the lifecycle: onboarding, adoption, retention, expansion, feedback. And it does it in inbox, in-app, chat, or a live-style presentation, in 17 languages, without the vendor lifting a finger per account.

Why the math matters

The reason this framing lands is the shape of the problem. A good CSM can carry maybe a few dozen accounts with real attention. Past that, quality degrades in a straight line. The industry's answer has always been to hire - which is expensive, slow, and caps out anyway. Singh's answer is to decouple coverage from headcount entirely.

The claim only means something if customers see results, and the ones Cast.app puts forward are specific. Pure Storage reported roughly $1.6 million a year added to the bottom line and 30x return on the tool. Aruba saw renewals climb 6.1%. Across the platform, Cast.app says it delivers more than 200,000 AI-presented business reviews every month - the kind of volume no staffed team reaches.

The quiet revenue

There is a reason a customer success founder ends up here rather than in sales tooling. The loudest revenue in software is new logos. The quietest - and often the largest - is the revenue a company already has and slowly lets slip: the renewal nobody chased, the expansion nobody surfaced, the account that churned because it never really got onboarded. Retention is where SaaS economics live or die, and a surprising amount of it still runs on spreadsheets, goodwill, and whoever happens to have time.

More accounts shouldn't automatically mean more CSMs. Dickey Singh

That is the wager Cast.app is built on. Singh is not claiming the human CSM disappears - he is claiming the human should be freed for the handful of conversations that genuinely need judgment, while the software covers the long tail that used to get nothing. It is a contrarian position in a market busy shipping copilots, and Singh seems comfortable there. He made the case in depth on UpdateAI's [Un]Churned podcast, where the throughline was scaling success without expanding the org chart.

Whether autopilot becomes the default or stays a strong niche, the underlying observation is hard to argue with. Customer success was designed for a world where attention had to be rationed. Singh's bet is that it no longer does - and that the companies figuring that out first will keep revenue the rest quietly leak.

Questions people ask