The Problem

R&D teams face exploding data complexity, driving wasted effort, rising costs, and slower discovery.

• 75-90% failure rate: AI deployments fail in specialized biological domains
• 400% growth: Bioinformatics services to 4× by 2035, signaling unmet tooling needs and complexity
• $250B/yr lost: Biotech loses ~$250B annually to fragmented data and infrastructure debt
• 80% wasted time: Data scientists spend most time on data plumbing instead of discovery

The Solution

AI platform automating design and execution of data engineering pipelines and statistical experimentation.

• 100% deployment success: Purpose-built AI agents with domain-specific intelligence via context engineering
• Radical simplicity: Tools and integrations simplify complex biological data workflows
• 80% productivity gain: Autonomous workflows handle the toughest data work end-to-end
• Agentic Data Engineering: Automates ingestion → cleaning/transformation → lineage → orchestration across your R&D stack
• Context-Aware Reasoning: Domain-specific intelligence through advanced tool use and context engineering
• Provenance & Reproducibility: Versioned datasets, transformation lineage, and auditable workflows captured in a provenance graph
• Self-Driving Semantic Layer: Autonomously defines and centralizes research metrics, experimental KPIs, and statistical calculations

Product: VC-GPT

Our first product is VC-GPT, a Virtual Cells analytics agent built on research published at ICML'25.

• Domain-specific reasoning via in-context learning with standardized, expert-designed tools and APIs
• Autonomous and reproducible workflows for complex biological data analysis
• Bioengineering hardware stack API integrations for seamless lab connectivity
• Planner-Executor-Critic architecture: Breaks objectives into reliable multi-step workflows

Market Opportunity

• TAM: $500B AI in R&D
• SAM: $250B AI in physical sciences and engineering
• SOM: $200M ARR potential by 2034
• Beachhead: $25B AI in biotechnology
• Secondary Beachhead: $80B Healthcare Big Data analytics

Traction

• ICML'25 publication for foundational technology (funded by ICML, Harvard, and Chan-Zuckerberg)
• 5 businesses on waitlist
• 3 pilot interest forms from major biopharma company and startups
• Handshake revenue share agreement with MIT startup
• 2+ computational biologists interested in collaboration and testing product
• Antler VC residency, calculus.house at The Residency network
• Top 5 of 2000+ selected for Choose Good Quests pitch by Savant
• Live prototype "VC-GPT" demoed at calculus.house

All-MIT CS Founding Team

• Alex Velez-Arce - Founder & CEO/CTO: AI research & strategy at FAANG+, MIT CS, Harvard BioAI
• Jesus Caraballo - Founding Engineer: AI research HMS, backend engineering, MIT Computer Science
• built data & ml products accounting for $100M+ in revenue
• Open-sourced virtual cells AI platform with 30K+ MAU
• ICML-and-NeurIPS-published researcher in BioML
• Formerly SWE at Pinterest and AI researcher at Harvard