Aanal Shah is the founder and CEO of Zillion Infotech, which helps startups and growing businesses build custom web and mobile applications, SaaS platforms and AI-powered solutions, from initial design through deployment and ongoing support. Shah tells Laurent Cochet why the company starts with the business problem before choosing the technology, when automation beats AI, and what founders should ask before hiring a technology partner.
A father's advice and a gap in the market
Shah graduated in 2017 and started a job. On the first day, Shah's father, who was already running a business, gave clear advice: do not work a job your whole life, but work for a while to understand how business runs, how clients are handled and how decisions are made, then start something of your own. Shah worked for five years, and the father kept pushing whenever the job felt comfortable.
While working, Shah noticed a gap. Small businesses and startups would take a business problem to a tech company and come back with a solution that was too complex or too expensive for their stage, or a vendor would build exactly what was in the spec without asking whether the spec was right. Big companies have consultants and internal teams to bridge that gap; smaller firms usually do not. When Shah started Zillion Infotech in 2022, the idea was simple: understand the business first, then choose the technology.
Questions before solutions
When a client asks for new software or wants to add AI, the team starts with questions: what is the actual problem, what does it cost today, and what does success look like six months from now. For existing platforms, Shah advises releasing features one by one rather than five at a time, so the team gets real feedback from users, who sometimes see features differently from the client.
Shah's example is a SaaS platform in the US that helps consumers find class action settlements they are eligible for and file the claims. The client wanted AI, but the claim forms had fixed fields and the user data was already in the system. The smart answer, Shah says, was automation: check whether the claim is signed and the subscription is active, then generate completed claim forms in bulk with one click.
The company started with e-commerce work and has since worked in healthcare, finance and travel tech. Shah calls healthcare and fintech its strongest areas.
Architecture comes before speed
For Shah, building is not the hardest part, because developers can now build quickly with AI. The harder part is the architecture behind the code. Shah could promise to build something in two days, but if the architecture is not decided on day one, scaling from 100 users to 1000 or 100k may mean rewriting the code from scratch. So instead of promising an MVP in five days, the team first decides on the architecture and asks about the client's vision and goals.
Once a product is live, Shah balances fixing existing issues with adding new features, and writes code so that changes do not affect current users. Running the company has also meant a different kind of responsibility: in a job, Shah worked from 10 to 7, but as a founder may need to answer a client at 3 am. The flexibility is good, Shah says, but the responsibility is infinite.
When AI is worth the cost
Shah sees every client asking for AI, but argues it should not be added by force. A chatbot may help a healthcare business, while a manufacturer that mainly showcases its portfolio and generates leads may get little from one. AI also costs more to run, because it uses tokens, and tokens come to money in the end.
Before a business spends on AI, Shah would ask whether it is beneficial in terms of money, time and effort. If a manual task takes 8 hours and AI cuts it to 1 or 2 hours, the gain is clear; adding AI only to follow the market trend is not a reason. An experiment is for when the outcome is unknown, Shah says, and AI belongs in a real project only when the team is sure it will benefit the end user.
Growing through retention and partnership
For the next 18 months, Shah defines success first as retaining existing clients, since losing them would signal a problem. The second goal is to train the current team of 15 people rather than hire more developers, so that with AI they can produce the output of a hundred. The third is long-term relationships in which Zillion Infotech acts as a technology partner, not just a vendor.
Shah prefers depth to breadth: rather than chase hundreds of domains, the aim is 10 good client relationships, for now centred on healthcare, fintech and e-commerce, while still learning new domains along the way. Given unlimited resources, Shah would automate processes so that the business no longer depends on the founder, freeing time to focus on other things.
Shah's advice to founders hiring a technology partner is to study its process. A provider that simply tries to close the deal is working for itself. A provider that tries to understand the problem in detail, asking many questions about why, when and what, is trying to solve the client's problem.
Watch the full conversation
Watch the full conversation above. Shah invites listeners to reach out through the Zillion Infotech website or by email to discuss their problem first.
AI InterConnect hosts conversations with founders, investors and operators building with AI. Explore more episodes on the AI InterConnect podcast page.