Pranay Dinavahi

CTO @hakaseAI_
Transforming the pharma value chain with HAI-Ops powered BioDigital Twins.

[Pranay Dinavahi @Pranay_Dinavahi]

@HakaseAI_ Exactly! You can have great tools individually, but if the researcher still has to keep jumping between systems and piecing the context back together, the workflow is still broken.
That’s exactly the gap we’re trying to close.

[Pranay Dinavahi @Pranay_Dinavahi]

@EricTopol This is the balance I like. Be ambitious about AI, but humble about biology. We can accelerate medicine massively without pretending every biological problem becomes a software problem.

[Pranay Dinavahi @Pranay_Dinavahi]

@DarioAmodei I share the optimism, The biggest opportunity may not just be finding better drugs, but making fewer bad decisions between discovery and the clinic. That alone could change the economics of drug development.

[Pranay Dinavahi @Pranay_Dinavahi]

@rohanpaul_ai This combination is powerful: serious compute, proprietary experimental data and a tight connection to the wet lab. AI improves quickly when its predictions meet biology quickly—and biology is allowed to disagree.

[Pranay Dinavahi @Pranay_Dinavahi]

@TimesOfAI_ Protein design is such a beautiful example of AI expanding human creativity. The goal isn’t to make biological intuition irrelevant—it’s to let researchers explore possibilities that intuition alone may never have reached.

[Pranay Dinavahi @Pranay_Dinavahi]

Best thing about a small team: the person who found the bug, the person fixing it, and the person deciding whether it matters are often the same three people in one conversation. I'll miss that when we're bigger at @HakaseAI_ 😅

[Pranay Dinavahi @Pranay_Dinavahi]

@yaireinhorn @TheEconomist @TimOpler @bradloncar @SamFazeli8 @BiotechCH @EliLillyandCo @BiopharmIQ @Maximus_Holla @Pharmdca Transformation at this scale rarely comes from one dramatic decision. It comes from years of disciplined choices while everyone else is watching the quarter. Technology becomes an advantage only when leadership turns it into a lasting capability.

[Pranay Dinavahi @Pranay_Dinavahi]

@Shea_ARK China’s progress is a reminder that scientific talent alone isn’t enough. Speed comes from connecting discovery, capital, clinical execution and decision-making without losing years at every handoff.

[Pranay Dinavahi @Pranay_Dinavahi]

“Your data never trains our models” sounds reassuring.
My next question: what would have to fail for that to become false?

If a config change is enough, privacy still depends on a promise.

At @HakaseAI_ , we want architecture, not policy, to enforce that boundary.

[Pranay Dinavahi @Pranay_Dinavahi]

@NVIDIAHealth This is exciting because scientific agents need real tools, not just better conversation. Connecting structure, chemistry and biological evidence while keeping the scientist firmly in control feels like the right direction!

[Pranay Dinavahi @Pranay_Dinavahi]

@johncumbers @AnthropicAI @SynBioBeta @Xaira_Thera Right! there’s a huge distance between answering a biology question and helping a team run a drug program. Provenance, uncertainty, failed experiments and accountable decisions live in that distance, and honestly, that’s where the interesting work is.

[Pranay Dinavahi @Pranay_Dinavahi]

@zbruceli We’re getting remarkably good at designing promising molecules. The humbling part is that biology still has a very long interview process. The teams that connect design, validation and clinical translation will create the lasting value here🤔

[Pranay Dinavahi @Pranay_Dinavahi]

@DaphneKoller This is the distinction the industry needs. AI can compress search and improve decisions, but biology still gets the final vote. The real breakthrough is building systems that expose uncertainty early, before teams commit years and millions to the wrong hypothesis.

[Pranay Dinavahi @Pranay_Dinavahi]

AI enabled assets in human trials. 60 have finished Phase 1. Eight have finished Phase 2. Humbling and
remarkable at the same time, and I think we should keep saying both.

[Pranay Dinavahi @Pranay_Dinavahi]

@RegXtrakt RegXtrakt really said, "Everyone line up, one source of truth only."
Refreshing to see the conversation shift from more dashboards to better architecture 🫡

[Pranay Dinavahi @Pranay_Dinavahi]

Drug discovery is basically Tinder for molecules.

Millions of candidates.
Everyone looks promising at first.

Then the red flags start showing up.
That's why preclinical triage matters.

We're building Hakase AI to make those red flags easier to spot.
https://t.co/GdqxgdtPNc

[Pranay Dinavahi @Pranay_Dinavahi]

Exciting, AI in drug development is advancing at an amazing rate. We've been building something to help accelerate drug development using AI as well 👀

[Nagma parween @TechByNagma]

INSTEAD OF WATCHING AN HOUR OF NETFLIX TONIGHT.

This 60-minute Cambridge lecture by Demis Hassabis will teach you more about the future of AI than most people will learn in the next 5 years.

Bookmark it and give it an hour, no matter what.

[Pranay Dinavahi @Pranay_Dinavahi]

Drug development is one of humanity's hardest engineering problems.

I've spent the last few months exploring how Al can help researchers make better decisions before expensive lab experiments begin.

Excited to share what we've been building.
#drugdiscovery