Predicting cancer drug responses before clinical trials begin

This article is a part of Startup Spotlight, a series that features young, up-and-coming startups.

Image credit: Timmy Loen

Decades of watching oncology drugs move from lab bench to clinical trials revealed a glaring flaw. Strong compounds regularly fail because they are tested on the wrong patients.

AlphaGrid tests every drug-patient pairing virtually before a trial begins. Its long-term goal is to test every drug-patient pairing virtually, before a single patient is enrolled.

😟 Problem

Over 70% of Phase II oncology trials fail due to efficacy issues from mismatched patients. This disconnect causes an estimated US$50 billion to US$60 billion in wasted trial spending annually. Current testing methods struggle to identify appropriate trial participants.

💡 Solution

AlphaGrid builds digital twins of patient tumors to predict specific drug responses. The platform uses a proprietary biological model trained on real organoid data. 

The startup’s biological discovery engine offers several key features, including:

• It converts standard US$5 pathology slides into queryable digital twin embeddings.
• The engine outputs spatial cell maps and per-drug response probabilities.
• Pharmaceutical companies can computationally screen molecule-indication pairs to stratify patient cohorts.

Image credit: AlphaGrid

📊 Market size

Oncology accounts for 48% of all AI drug discovery spending and 41% of clinical trials. The global cancer diagnostics market represents a US$158 billion opportunity. The broader AI drug discovery sector is projected to reach US$160 billion by 2035.

🤝 Team

• David Pyrce. Executive chair. A biotech entrepreneur with 30 years in oncology drug development, leading four IPOs and securing over US$5 billion across transactions.
• Ananta Balaji. CTO. A former research scientist at Nokia Bell Labs Cambridge with five patents and 10 publications.
• Sachin Yaragal. CEO. Previously founded Ardeo and led product and growth across early-stage and scaled startups.

🚀 Traction

• Signed a statement of work with a premier oncology research institution in India.
• Ingested over 500 million biological data points to generate live data from ovarian cancer cases.
• Established a partnership with a large medical center in India and launched a live hospital pilot.
• Entering active discussions with potential pharmaceutical partners and academic medical centers worldwide.

🏆 Competition

AlphaGrid competes with AI-native cancer biology models like Noetik and Pathos AI. Other players focus on descriptive biology or cell-level simulations. AlphaGrid uses patient-derived organoids to preserve actual drug response behavior and microenvironment interactions.

💰 Financials

🚩 Risks