Technosport CTO on why scaling a business is rarely a server problem
At DevSparks Hyderabad, Achal Sharma, CTO of Technosport, said scaling a fast-growing consumer brand depends less on adding infrastructure and more on identifying where the true constraints lie.
Engineering teams tend to judge scale by how well systems perform: whether uptime holds, servers keep up, releases ship on time. For a brand that sells across D2C, mobile, and retail at once, that measure only tells half the story. The harder test is whether technology choices track the business itself.
That was the question at the heart of a lightning talk at DevSparks Hyderabad 2026, titled ‘Beyond Code: Building Technology That Scales Businesses.'
The event brought together developers, technology leaders, startups, and Global Capability Centres for sessions spanning agentic AI, cybersecurity, cloud infrastructure, and developer productivity.
The talk was delivered by Achal Sharma, Chief Technology Officer at Technosport, the Bengaluru-based performance wear brand that manufactures in-house, from yarn to finished garment, alongside running its own retail and digital business.
Sharma brings close to two decades of engineering leadership to that question, with stints at Myntra, Mobile Premier League, and Wakefit before joining Technosport.
When the bottleneck isn't the server
Sharma pointed to a recurring assumption in his own career: that a traffic surge during a sale is solved by adding servers. "It's not always the servers, the architecture, or a well-defined system that's the culprit," he said.
He described instances where server capacity held up fine during traffic surges, but the actual bottleneck turned up elsewhere: inventory availability, code velocity, or another part of the product experience entirely. The fix, he said, depends on correctly identifying where the failure lies, not defaulting to infrastructure as the answer.
Forecasting demand, fixing gaps
Sharma extended the point to customer experience. Users don't talk about checkout APIs or backend architecture. They notice whether an order arrived on time and what went wrong when it didn't.
At Technosport, that experience runs through a layer most digital-first brands skip: manufacturing, sizing, and store inventory. The company uses AI for demand forecasting alongside efforts to close gaps around inventory accuracy, store availability, and product-image accuracy.
He said Technosport closed the last financial year at Rs 600 crore in revenue, is targeting more than Rs 1,000 crore this year, and is moving its website from Shopify to an in-house build.
Should we build this at all?
On the use of AI tools in engineering workflows, Sharma stressed that specificity matters more than the tool itself. "If you summarize your requirement from a business perspective and give that to the agent, it will give you a better result," he said, contrasting this with vague prompts that produce generic output.
He extended that logic to how engineers evaluate their own work, arguing that technical feasibility should not be the only filter for what gets built.
"Can we build this? The next question is: should we build this?" he said.
The throughline of Sharma’s session was that engineering decisions carry more weight when made with business context rather than in isolation. Infrastructure choices, AI adoption, and even the debate over what to build at all, he said, are extensions of the same discipline: understanding what a business actually needs before deciding how to build it.
Edited by Teja Lele