A joint Google and Accel accelerator sifted through more than 4,000 applications from Indian AI startups and rejected roughly 70% outright for building little more than chatbot layers on top of existing software — then selected five companies doing something harder.
The five winners of the latest cohort of the Atoms program will each receive up to $2 million in funding from Accel and Google’s AI Futures Fund, plus up to $350,000 in cloud and compute credits from Google, according to the announcement.
Accel partner Prayank Swaroop said the rejected “wrapper” applications shared a common flaw: they layered AI features onto existing software without “reimagining new workflows using AI.” Most of what remained — applications that cleared that bar but still didn’t make the cut — came from crowded sectors. Marketing automation and AI recruitment tools dominated, areas where investors found little to distinguish one entrant from another.
Who Got In
The five selected startups are: K-Dense, building an AI co-scientist for life sciences and chemistry research; Dodge.ai, developing autonomous agents for enterprise ERP systems; Persistence Labs, focused on voice AI for call center operations; Zingroll, a platform for AI-generated films and shows; and LevelPlane, applying AI to industrial automation in automotive and aerospace manufacturing.
Jonathan Silber, co-founder and director of Google’s AI Futures Fund, said the selected companies align with areas where AI is expected to reach deeper real-world adoption. The program does not require startups to use Google‘s models exclusively — many combine multiple models depending on the task, he noted.
The arrangement has a deliberate feedback loop built in. Insights from how startups deploy models in production flow back to Google DeepMind teams. Silber called it a “flywheel” between startup experimentation and model development. “If a company is using an alternative model, that means Google has work to do to build the best model in the market,” he said.
What the Applications Revealed
The volume of submissions — nearly four times that of previous Atoms cohorts — reflected the surge of first-time founders entering India’s AI ecosystem. The composition also revealed where that ecosystem is concentrated: about 62% of submissions focused on productivity tools, and another 13% on software development and coding. Consumer products were a small fraction of the total.
Swaroop said he had hoped to see more ideas in healthcare and education. Neither sector featured prominently in the applicant pool.
The Atoms program was announced in November as an early-stage vehicle specifically for AI startups building products tied to India.
Photo by RDNE Stock project on Pexels
This article is a curated summary based on third-party sources. Source: Read the original article
