By Dr. Sunil Shekhawat, Founder & CEO, SanchiConnect
India’s deep tech is attracting significant capital, policy attention, and infrastructure at an ever-growing scale. Indian DeepTech companies raised about $2.9 billion across 189 PE-VC deals in 2025, while cumulative funding between 2015 and July 2026 reached $11.4 billion. The deeptech sector’s share of India’s PE-VC activity has also risen from around 4% in 2016 to 15% in 2025. By July 16, 2026, DeepTech had already recorded 103 deals during the year.
Policy is also being designed around longer technology cycles. In February 2026, the government extended the recognition period for DeepTech startups from 10 to 20 years and raised the turnover ceiling to ₹300 crore. The ₹1 lakh crore Research, Development and Innovation Fund is also intended to support high-risk, high-impact R&D. But funding a DeepTech company and commercialising its technology are two different challenges, and this is where enterprise procurement can become an unexpected bottleneck.
Procurement is built around risk
For a large enterprise or public institution, procurement is designed to reduce uncertainty. A typical Source-to-Pay (S2P) process can move through vendor empanelment, RFP/RFQ, technical and legal evaluation, financial checks, contracting, purchase orders, delivery validation, and three-way matching between the order, receipt, and invoice. There is a reason for this methodical process. Buyers need to protect budgets, maintain audit trails, and ensure that a supplier can actually deliver.
DeepTech complicates that equation because the buyer may be assessing something for which there is no established benchmark. The difficulty emerges when the technology itself is new. An advanced AI system, semiconductor design, industrial robotics platform, drone, or quantum solution may have few comparable suppliers, limited deployment history, and no established benchmark for price or performance. It can be whether the buyer is comfortable becoming an early customer for a technology category that is still taking shape.
DeepTech does not always fit the vendor checklist
A startup may have developed a technically validated product but have limited commercial revenue. It may have strong intellectual property and engineering capability without the years of enterprise deployments that conventional procurement criteria often favour. This is where conventional requirements around prior experience, turnover, financial capacity, and vendor history can become difficult for young DeepTech companies.
India has already recognised part of this problem. Through GeM Startup Runway, eligible DPIIT-recognised startups can receive exemptions from prior experience, prior turnover, and earnest money deposit requirements in specified public procurement. There are about 37,758 startups registered on GeM. That is meaningful because it shows that access to government demand is no longer entirely tied to being an established vendor. But the bigger question is what happens after the startup gets through the door. For a DeepTech company, listing on a procurement platform does not automatically solve technical validation, integration, risk assessment, or the buyer’s hesitation to deploy something new. So the challenge has shifted from access to procurement to conversion of access into adoption.
The delay does not stop the startup’s costs
NITI Aayog says that TRL 4 to 6 is DeepTech’s “Valley of Death”, the stage where technologies can struggle to move from development towards commercialisation and scale. This is where procurement timelines matter most. A DeepTech company can spend heavily on specialised engineers, hardware, testing, compute, certification and product development before recurring revenue begins. Every additional month spent waiting for a commercial decision extends that gap.
This is where procurement timelines matter most. A DeepTech company can spend heavily on specialised engineers, hardware, testing, compute, certification and product development before recurring revenue begins. Every additional month spent waiting for a commercial decision extends that gap. It also creates a funding problem. Investors can fund technical development, but a signed enterprise contract provides evidence that someone is willing to pay for the technology at scale. Without that evidence, the next funding round can become harder just as the company’s cash requirements increase.
Pilots can create the missing evidence
A paid proof-of-concept, controlled innovation sandbox or milestone-based pilot gives an enterprise a way to test performance without immediately committing to large-scale deployment. For the startup, it creates something equally valuable: real-world data, a customer reference and evidence that can support subsequent procurement and fundraising.
Co-development can take this one step further, particularly where the technology needs access to enterprise data, infrastructure or specialised operating environments. But these arrangements also need clear boundaries around IP. A DeepTech startup’s core intellectual property can underpin its valuation and future licensing opportunities; the right to use a solution should not automatically become ownership of the underlying technology.
India has built the capacity, so demand must follow
The semiconductor sector shows how quickly India’s technology ambitions are becoming capital-intensive. There were 12 semiconductor projects involving more than ₹1.64 lakh crore that had been approved this year, while Semicon 2.0 carries a ₹1,27,500 crore outlay. Twenty-four semiconductor design projects from startups and MSMEs have received support, and 105 startups and MSMEs have received access to EDA tools.
But investment in capacity does not automatically create commercial demand. A chip still has to be qualified by a customer. An AI system has to be deployed inside an organisation. A robotics platform has to prove itself on the factory floor. A quantum technology has to move from demonstration to an actual use case. That is why procurement deserves to be treated as part of India’s DeepTech commercialisation strategy. The country has spent years building the capital, infrastructure and policy support needed to create new technologies. The next challenge is making sure those technologies can find their first customer, first deployment and first repeat order by focusing on identifying the real gap that lies between technology that works and technology that gets bought, deployed, and scaled.