Investors are increasingly finding it difficult to separate genuine AI innovation from hype, as many startups use AI as a buzzword
While some focus on evaluating founders, problem understanding and team-building capabilities at the pre-revenue stage, others stress on AI model accuracy, compute cost and sustainable margins
AI startups are now hiring senior engineering and product leaders much earlier to improve operational efficiency, manage rising compute costs and maintain healthy margins while scaling
India’s native AI startups have attracted $1.8 Bn+ in funding since 2020. But 86% of funding is concentrated in applications rather than infrastructure and foundation layers. While this has accelerated commercialisation, it has also raised questions such as what investors actually fund in AI today, what an investor-ready AI data room needs, and what the emerging red flags in the space look like.
“Investors today struggle to separate genuine AI innovation from hype, as many startups globally use AI as a buzzword without demonstrating real technical advancement,” said Archana Jahagirdar, founder & managing partner at Rukam Capital.
Jahagirdar was speaking at the third edition of Inc42’s AI Summit in Bengaluru during a panel discussion titled “What AI Investors Look For & What Kills Deals Before IC”. She was joined by Ashwin Raguraman, cofounder and partner at Bharat Innovation Fund, while the session was moderated by Bala Srinivasa, managing director at Arkam Ventures.
Jahagirdar noted that building AI products significantly increases costs, making it critical for investors to assess whether AI truly creates meaningful differentiation and justifies the added expense.
Raguraman added that the bigger challenge now lies in evaluating customer adoption of AI solutions, especially in B2B startups. He explained that even strong enterprise logos and pilot projects no longer guarantee credibility, as companies may still be experimenting with AI rather than committing mainstream budgets, making it harder for investors to assess go-to-market strategies and identify real customers.
The panellists also have a different take on key investment metrics they look at while investing. Jahagirdar invests primarily at a pre-revenue stage, and the focus is still largely on evaluating the founder’s understanding of the problem and their ability to build the right solution and team over time.
On the other hand, Raguraman is keen to understand how startup founders explain improvements in AI model accuracy, management of compute costs, and the ability to build sustainable margins while scaling engineering and product capabilities at an early stage.
“Gross margin is still very important for us as a deep tech fund. I think we like to see anything north of 60% when it’s software, and if it’s hardware, maybe 40%, but we like to see those numbers still,” he added.
He added that AI startups are now prioritising senior engineering and product leadership much earlier than before, driven by rising compute costs and the need for operational efficiency. He noted that bringing in experienced technical leadership early helps startups maintain competitive margins while scaling their AI capabilities.
Source: Inc42 - Startups
