A senior engineering leader leaving a fourteen-year tenure at a major tech giant always draws attention. Manohar Paluri fits this description perfectly.
The IIIT Hyderabad alumnus steadily rose from a computer vision researcher to Meta’s Vice President of Artificial Intelligence. Now, he has co-founded Warp Sciences. Interestingly, this bold new venture recently secured a massive $215 million in funding.
Rather than building another consumer app a common path for founders exploring the limits of specialized vs. generalist AI Paluri is redirecting his deep expertise toward a stubbornly physical problem: the agonizingly slow pace of discovering and manufacturing new medicines.
Warp Sciences aims to fundamentally rewire pharmaceutical operations by fusing artificial intelligence, cutting-edge robotics, and traditional chemistry into a single cohesive system.
The Physical Bottleneck in Pharmaceutical Development
To understand this approach, we must examine where modern drug development actually stalls. In recent years, computational biology has made massive strides. Highly sophisticated software models can now predict protein folding.
Furthermore, they show how specific molecules might bind to a target. Consequently, a skilled computational chemist can design a new therapeutic compound digitally in a single afternoon. However, physically synthesizing that digital concept in a laboratory often takes weeks of painstaking manual effort.
Every subsequent round of testing and refinement is forced to wait on this physical chemistry bottleneck. According to an India today, detailed industry profile on his career transition Paluri saw an opportunity to apply artificial intelligence to a vastly different domain directly impacting human health.
Currently, pharmaceutical companies offshore much of this tedious synthesis work. A typical drug program requires chemists to manually refine thousands of variations during lead optimization. Unfortunately, the problem goes beyond slow human hands.
The collective learnings from these isolated manual experiments simply do not automatically compound. As a result, they fail to improve the broader manufacturing process. In short, the industry runs high-tech software on analog, fragmented hardware.
Automating the Lab and Learning from Failure
Warp Sciences operates as an advanced contract research organization that brings this critical infrastructure back to domestic soil through heavy automation. The core mechanism is a continuous loop between software and hardware.
When a pharmaceutical client needs a batch of drug candidates synthesized and tested, Warp’s artificial intelligence tools first help plan the most efficient chemical route to create each compound.
From there, robotic systems take over the highly repetitive, high-volume steps of the physical process, handling the precise measuring, setting up of reactions, and purifying of chemical samples.
This approach highlights exactly how to build a defensible AI system that scales reliably from a digital prototype into real-world production. By offloading mechanical execution, expert human chemists are freed up to focus entirely on the hardest molecular problems.
The true operational advantage of Warp’s system lies in how it handles data collection, specifically regarding failure.
Traditional published chemistry data disproportionately records only the experiments that worked. Countless failed reactions and dead ends are simply discarded or locked away in isolated corporate notebooks.
Because Warp operates its own robotic fleets to run every single physical experiment, the system automatically logs every meticulous step, including the exact environmental conditions of the reactions that fail.
As the robots execute more chemistry, the underlying models become increasingly accurate at predicting the best synthesis pathways. Every single experiment makes the next one smarter, effectively turning the wet lab into a continuously optimizing physical search engine for life-saving medicines.
