Data Science Ventures

Data Science Ventures

Data Science Ventures was founded in 2016 by Srinivas Akkaraju, a former Genentech executive and healthcare investor, alongside Naveen Bhatia. The firm emerged from a thesis that machine learning could systematically identify high-potential technical founders earlier than conventional networks. Rather than relying on warm introductions and geographic concentration, DSV built a data ingestion engine that analyzes patent filings, academic publications, technical hiring patterns, and product launch trajectories — a scouting infrastructure unusual for a firm of its vintage and size. The firm invests primarily at the Seed and Series A stages, with check sizes typically ranging from $1 million to $5 million. Its strategy spans enterprise software, AI/ML infrastructure, digital health, fintech, and cybersecurity. DSV reserves significant capital for follow-on investments in portfolio companies that demonstrate technical milestones and product-market traction. Public portfolio exposure includes positions in AI-native drug discovery platforms, machine learning operations tooling, and precision medicine diagnostics. Geographic coverage leans heavily on US-based startups, with selective co-investments in Canadian and UK technical hubs where deep talent pools intersect with strong IP regimes. DSV maintains a lean partnership structure with Akkaraju overseeing life sciences and digital health allocation while Bhatia leads enterprise and infrastructure investing. The firm has not publicly disclosed total AUM or headcount, consistent with its posture as a tightly held early-stage specialist. In March 2024, DSV participated in a Series A round for a computational biology startup applying transformer models to protein design (per Crunchbase, 2024). The firm occasionally syndicates with more generalist, multi-stage funds when portfolio companies require larger growth rounds, but its origination remains independent. What structurally differentiates DSV is its sourcing model — an algorithmic discovery layer that operates continuously, screening for technical founders before they actively fundraise. This quantitative top-of-funnel approach sidesteps the relationship-dependency that characterizes most Seed-stage venture firms. The architecture lets a small partnership maintain a high-volume, technically filtered deal pipeline without scaling headcount proportionally.