Tattvam AI has emerged from stealth with $1.7 million in pre-seed funding led by Seedcamp. Other participants include EWOR, Entropy Industrial Ventures, Concept Ventures, and semiconductor industry veteran Stan Boland.
The company is building what it describes as an “AI layer” for chip design — a system intended to automate complex semiconductor development tasks that today require highly specialized engineers and years of iteration.
The Push Toward Custom Silicon
Demand for custom silicon has surged as artificial intelligence workloads grow larger and more specialized.
Unlike general-purpose processors, custom chips are optimized for specific tasks such as AI training or inference. These application-specific processors can deliver significantly higher performance and improved energy efficiency compared to conventional hardware.
Major technology firms are already investing heavily in custom designs. Google has developed Tensor Processing Units (TPUs) for AI workloads, while NVIDIA has expanded into specialized AI inference hardware partnerships. UK startups including Fractile are also pursuing application-specific architectures.
As AI models scale and industries from autonomous vehicles to drug discovery demand more compute, custom silicon has become a competitive differentiator.
A Bottleneck in Design
Despite the urgency, chip design remains a slow and resource-intensive process.
Developing a new processor typically involves:
- Circuit architecture design
- Verification and simulation
- Physical layout optimization
- Iterative performance tuning
The process can take two to three years and relies on a relatively small pool of experienced semiconductor engineers.
Meanwhile, on the software side, AI systems are already generating complex code and assisting in high-level reasoning tasks. Tattvam AI’s premise is that similar advances can be applied to hardware design.
Reasoning-Based AI for Circuits
Tattvam AI is building a system designed to understand circuit structure and autonomously solve complex design problems.
“Chip design is fundamentally a reasoning problem over an enormous search space,” said Bragadeesh Suresh Babu, CEO and co-founder of Tattvam AI. He argues that existing large language models lack the deep structural understanding required for circuit-level reasoning.
Instead of relying solely on pattern recognition, the company is developing models intended to reason over constraints, trade-offs, and interdependencies in chip architectures — mimicking how experienced engineers approach design challenges.
The goal is to automate key portions of the electronic design automation (EDA) workflow, dramatically reducing development timelines.
Founder Background and Technical Roots
Bragadeesh is an alumnus of Indian Institute of Technology Madras (IIT Madras), widely regarded as one of India’s top engineering institutions.
He previously worked as an early engineer at UK-based brain-monitoring startup CoMind and later joined Fractile, a chip startup focused on AI processors. He turned down an offer to join Google’s TPU team to launch Tattvam AI.
He co-founded the company with Lannan Jiang, who has been developing chips at a research lab at ETH Zurich.
Stan Boland, a semiconductor veteran who previously founded Icera and Element 14, said the startup’s approach could dramatically speed up the iterative process of chip design using EDA tools.
Why This Matters Now
Semiconductor development sits at the intersection of national industrial strategy and AI competitiveness.
Governments in the U.S., Europe, and Asia are investing heavily in chip manufacturing and design capacity. However, scaling physical fabrication capacity alone does not address design bottlenecks.
If Tattvam AI can meaningfully shorten chip development cycles, it could:
- Lower barriers for startups building custom processors
- Reduce costs associated with iterative silicon tape-outs
- Enable faster experimentation in AI-specific architectures
In an industry where time-to-market can define competitive advantage, shaving months — or years — from development could reshape how companies approach silicon strategy.
Early Stage, High Stakes
At $1.7 million, the funding round is modest compared to the capital intensity of semiconductor ventures. However, Tattvam AI’s focus is on the design layer rather than fabrication.
The company plans to launch its first product in the coming months and is working with early partners to accelerate next-generation chip development.
The broader bet is clear: as AI transforms software engineering, similar reasoning systems may soon transform hardware engineering.
In a world increasingly defined by custom silicon, the automation of chip design itself could become the next frontier of AI.
