Thought Behind Things
Why AI startups are turning back to services
GTM lead Muhammad Qasim Shamim Ali Khan explains why building an AI product got easy while selling one got hard, and why some startups are adding services to survive.
Contents
The problem: building got easy, selling did not
The episode opens with a question Muzamil says he keeps hearing online: if anyone can build an AI product with Claude and agents, who will buy it? Building is now easy, but most people building right now have no network, no track record, and no sales experience. Muzamil says many of them are “completely clueless” about go-to-market, and that is why he wanted to talk about selling.
Guest Qasim Ali Khan, who has spent about eight years in technology sales and now works on the sales side of Ruby AI, agrees that the gap between building and selling is the central problem.
What Ruby AI does
Qasim describes Ruby as an AI go-to-market harness that collects company context, product context, and sales data, then turns it into deal intelligence for sales reps. New reps often take a couple of weeks to ramp up, and much of the useful knowledge sits in past calls and deals that nobody revisits. In his words, Ruby “learns itself from your prior deals, from your company, context,” then guides reps on what to do next: which message to send, which deck to use, and what to say.
He positions Ruby as a layer on top of existing tools rather than a replacement. It connects to CRMs such as Salesforce and HubSpot, video tools, and Gmail, so reps do not have to re-enter information. Qasim says Ruby is approved on the Salesforce marketplace after completing the required steps.
Why SaaS feels “dead” to many founders
Muzamil raises the familiar argument that SaaS is dying because any company can build what it needs in-house. Qasim’s answer is more about buyer behaviour. When the market is flooded with similar tools, customers get fatigued: “I just don’t have time to go and analyse all of them,” he says, so they often decide not to buy at all.
He compares it to a market where fifteen shops sell the same t-shirt and the price collapses. The software itself has become commoditised. In his experience, what sells more easily is outcomes rather than software. He describes a case where a tool that costs a customer $300 in software can deliver a $3,000 outcome, with humans still in the loop. Muzamil connects this to the idea of “service as software,” which he calls the next big company in the field, and Qasim agrees that this is the direction he sees.
Ruby’s pivot toward services
The most concrete part of the conversation is Ruby’s own pivot. Qasim says the team has been running a normal outbound motion for a couple of months, with free trials, and the response was cold: “hardly a few people sign up.” What worked was co-founder Chris’s network. Chris is a sales veteran who led sales at Salesforce in Canada for years, and people took calls because of the trust he carried.
That network also opened a second revenue stream. A ClickHouse-style database company was selling well but had no implementation team. Ruby’s engineers had experience with a tool the company used, so Ruby became their implementation partner. Qasim describes the result as “a bit of Chris’s network plus our technical expertise.” Ruby now has two revenue streams: the SaaS product, and implementation services for these customers.
He is clear that this is partly survival. “We need to get the cash flows to keep our team running,” he says. He also argues that the services work keeps the engineering team working with current tools, which can feed back into the product. He calls it a smart move, while acknowledging it is not easy: you need a solid reputation with a big partner, and you cannot afford to disappoint them.
He adds that many startups raised small friends-and-family rounds during the AI rush and are now confused about how to sell what they have built. For them, services are less an extension of their work than a lifeline while the team figures out distribution.
Outcome pricing and the exit question
Qasim describes how many North American founders think about exits. They often aim for early exits, building a product that solves one use case well and selling it to a larger company where it fits upstream or downstream. His example is personalised outreach, a feature a larger CRM could absorb. He says the founders he knows are looking less at a billion-dollar company and more at a 20 to 30 million dollar exit or an acqui-hire.
He is honest about the risk. HubSpot could copy the feature, and bigger companies may simply want to acquire the team. He also describes a customer who had seen a competitor struggle with an AI writing assistant and was glad to find Ruby doing it well.
What to build, and where
Muzamil pushes on whether Pakistan should move toward product development or toward deployed engineering services. Qasim notes that the tech scene has long been comfortable with services, with clear sales funnels and project pipelines, and that inertia makes it hard for those companies to retrain their teams toward products. Funding is the bigger issue: a product takes months to build, and with AI credits and tools costing a lot, a company without money cannot afford to burn its savings.
On the individual level, Qasim says he would personally choose product building, but only with a smart team, a globally scalable idea, and the right connections. Without those, he would advise engineers and freelancers to become forward-deployed engineers, because those skills stay relevant longer and let an individual move between companies. He returns to the episode’s central point: distribution and network often matter more than the product itself. Having a good idea is not enough if you do not know who your first customers are.
Staying steady in a turbulent market
Near the end, Muzamil raises the wave of tech layoffs, mentioning Microsoft’s announcement of cuts to around 6,000 of its workforce. Qasim says the market is slow and steady rather than a moonshot, and that “expectation management” matters. The old assumption that a single launch makes you a billionaire is probably gone.
He also reflects on his own career. He did not get the chance to work at a big corporation early on, and working with smaller companies has given him versatility. In a small company, the founder’s problem becomes your problem, which gives you a clearer view of the business. The trade-off is that people who have worked at large brands sometimes want that name on their CV afterwards.
His advice to young professionals in Pakistan is to work on more than one area at once, so that if one thing does not work out, you can pivot. Muzamil agrees the moment is turbulent, and the two end on a hopeful note that the sector can come out in a stronger shape.
Build with me in the Inner Circle.
You don't need to code. You need a problem you know well, and a room of people building beside you.

