Supreet
Deshpande

The Man Who Decided to Call Every Doctor Pharma Never Visits

The global pharmaceutical industry spends upward of $30 billion annually on field sales forces. It reaches roughly 30% of the physicians it targets. The remaining 70% - specialists in rural counties, overworked hospitalists, time-pressed oncologists - never get the call. Supreet Deshpande spent years inside one of the world's most influential consulting firms watching this problem repeat itself across every client deck. He stopped writing about it and started building the fix.

That fix is Synthio Labs: a San Francisco startup that deploys clinical-grade voice and chat AI as a kind of always-on engagement layer for life sciences companies. Where a traditional pharma rep costs $200,000 to $400,000 per year and can visit maybe a few hundred physicians, Synthio's AI agents can reach hundreds of thousands - speaking in multiple languages, citing FDA labels accurately, logging every insight back to headquarters in real time.

The company launched in 2024, cleared Y Combinator's Spring 2025 batch, landed on Forbes' list of the most promising startups from that cohort, and closed a $5 million seed round led by Elevation Capital in November 2025. Several of the world's Top 10 pharma companies are already running pilots.

From Data Science to Strategy to Builder

Deshpande's trajectory follows a kind of deliberate arc. He started in the numbers - quantitative analytics at ZS Associates, one of the most specialized consulting firms in the pharma-commercial space, where he worked on promotion response models, sales force sizing, and portfolio strategy. He passed through JPMorgan Chase. Then North Carolina State University, where he earned a master's in Advanced Analytics and, in 2019, won first place for "Best Insight" at the Carolina Data Challenge - a time series analysis of how the U.S.-China agricultural trade war was hitting sorghum farmers. That is not a typical portfolio piece, but it signals something: an instinct to dig into messy, politically complex data and surface something a general audience can actually use.

From NCSU he moved into McKinsey, where his focus eventually sharpened to "Gen AI for Life Sciences" - leading AI initiatives for the top 50 global pharma companies and biotechs across clinical development, medical affairs, and commercial operations. That work gave him an unusual double perspective: the strategic view of how pharma leadership thinks about AI, and the operational view of why most AI pilots die before production. He wasn't just reading about the problem. He was billing it.

The insight that became Synthio wasn't exotic. It was obvious, in the way that obvious things are only obvious after someone says them clearly. Pharmaceutical companies weren't failing to reach physicians because they lacked motivation. They were failing because the economics of human field teams make scale impossible. A voice AI agent that knows every word of an FDA label, never misquotes a contraindication, and can adjust its language based on whether it's talking to a cardiologist in Boston or a general practitioner in rural Texas - that's not a better rep. That's a fundamentally different category of engagement.

Building the Team

Deshpande co-founded Synthio with two former tech-company engineers whose backgrounds are conspicuously non-pharma. Sahitya Sridhar, the Chief Product Officer, built AI systems at Audible and Amazon - she led developer-efficiency Gen AI work and built scalable content personalization infrastructure. Rajashekar Vasantha, the CTO, comes from AWS AI research and holds an M.S. in Data Science from NYU. The configuration is intentional: one person who understands what pharma actually needs, two who know how to build systems that don't break under load.

The three founders share India-origin backgrounds and early careers built across both U.S. and Indian technology ecosystems, which shapes both the company's product thinking (multilingual from day one) and its operational footprint (engineering teams expanding across the U.S. and India simultaneously).

What Synthio Actually Does

Synthio's product suite is named like a constellation. Jarvis is a voice AI copilot for pharmaceutical field reps - not replacing them but augmenting them with real-time clinical reasoning and on-call FDA label data. Ather is the multimodal engine that handles physician and patient engagement across voice, chat, and web. Helix handles patient support with voice AI agents designed for direct-to-consumer healthcare brands. Polaris HQ is the data intelligence layer, turning every conversation into structured insight for headquarters.

Beneath all of it is what Synthio calls its Clinical Reasoning Engine - the infrastructure that makes compliance possible at scale. An AI that mispresents a drug's indication or misquotes a dosing guideline isn't just incorrect; it's a regulatory liability. Synthio's approach to this isn't a guardrails layer slapped over a generic large language model. It ingests FDA label data dynamically, ties it to each conversation, and keeps a verifiable audit trail.

An early proof of concept: a direct-to-consumer skin care brand used Synthio's voice AI for eczema patient engagement and fielded 5,000 calls in 48 hours. Industry benchmark response rates for human-staffed patient support programs: a fraction of that, at multiples of the cost.

Why Now, Why Them

The timing isn't coincidental. The convergence of reliable voice AI, production-ready large language models, and pharma's acute need to demonstrate ROI on field teams makes 2024-2025 the precise window when a product like Synthio becomes deployable. A year earlier, the voice quality wasn't there. Two years earlier, the clinical reasoning infrastructure would have required a team five times the size.

Elevation Capital led the $5 million seed round, with participation from 1984 Ventures, Peak XV Partners, and Y Combinator. The use of funds points to the next phase: deeper engineering in the U.S. and India, scaled enterprise deployments in North America and Europe, and expanded partnerships with global life sciences leaders.

Deshpande's stated ambition is for Synthio to become "the engagement layer for every pharma touchpoint" - a foundational layer, not a point solution. That's the kind of framing that makes investors attentive and incumbents nervous. The legacy CRMs, call-center infrastructure, and field force management platforms that currently sit between pharma brands and the physicians they need to reach are, by this logic, already obsolete. Synthio just hasn't finished replacing them yet.

For the physicians who have spent a decade avoiding sales calls, the company's pitch is different: this is an AI that shows up with the information they actually need, on their schedule, in their language. For the patients who needed answers about their medication at 2 a.m. and found a hold queue instead, Synthio is the call that finally comes through. That's the specificity Deshpande keeps returning to. Not "transforming healthcare." Reaching the 70% who were never in the plan.