NostrumAI — Neural Scaling for Drug Discovery
AI Drug Discovery · Neural Scaling · Autonomous Research

Applying neural
scaling laws to
living biology

NostrumAI is an established AI biotech company automating the full drug discovery cycle — from therapeutic candidate identification to autonomous laboratory experimentation — built by veterans of Genentech’s Prescient Design.

Our Platform The Science
💰
Annual Revenue
$50M USD
Founded
2020
Stage
Scaling
Domain
AI Biotech
Heritage
Genentech
$50MAnnual Revenue
2020Founded
AI+BioCore Discipline
The Platform

Built to close the loop

NostrumAI’s platform automates the complete design-run-interpret cycle for biological research — removing bottlenecks of manual intervention and replacing them with intelligent, self-directing workflows that scale with data.

01
🎯
Therapeutic Candidate Discovery
AI models that identify and rank novel therapeutic candidates across target classes — applying neural scaling to compress years of early discovery into accelerated, data-driven cycles.
02
🤖
Autonomous Research Engine
A fully automated design-run-interpret loop for biological experiments. The platform designs experiments, orchestrates execution, interprets results, and refines hypotheses — with minimal human intervention.
03
🧪
Preclinical Workflow Management
End-to-end management of complex preclinical workflows — from assay design to data integration — intelligently coordinated to reduce timelines and eliminate procedural overhead.
// The Autonomous Research Cycle
Design
AI generates experimental hypotheses and protocols
→
Run
Automated lab execution with no manual handoff
→
Interpret
Models analyse results and update the hypothesis space
↺
Iterate
Next cycle begins — smarter, faster, self-directed
The Science

Neural scaling, applied to biology

NostrumAI is built on a foundational scientific thesis: the scaling laws that transformed language models can be applied to biological problems — and the results will be similarly transformative.

Our models are trained on biological data at scale — protein sequences, assay results, molecular structures, genomic profiles — learning the deep statistical grammar of life science in a way that enables genuine generalisation to unseen targets and novel disease areas.

The team brings frontline expertise from Genentech’s Prescient Design group — one of the world’s most advanced AI-for-biology research organisations — combining pharmaceutical rigour with cutting-edge ML architecture.

The result is a platform that does not merely assist scientists — it reasons about biology at a depth and speed that no human workflow can match.

// Scientific Foundations
How NostrumAI thinks about biology
Neural Scaling Laws — The same principles that scaled language models are applied to biological foundation models, enabling emergent capabilities across drug targets.
Biological Data at Scale — Models trained on diverse, high-quality biological datasets: sequences, structures, assay results, and multi-omics profiles.
Closed-Loop Experimentation — Simulation and physical lab work integrated in a continuous feedback cycle — AI predictions validated and refined by real experimental data.
Preclinical Intelligence — AI-managed preclinical workflows that track, prioritise, and interpret complex multi-stage experimental programmes.
Leadership

Built by industry veterans

NostrumAI was founded by specialists who helped pioneer AI-for-biology at Genentech’s Prescient Design — one of the most respected computational drug discovery groups in the world.

S
Srikumar
CEO & CTO · Co-Founder
Founding technologist and executive leader of NostrumAI. Srikumar drives both the scientific vision and platform architecture — combining deep ML expertise with pharmaceutical domain knowledge built at Genentech’s Prescient Design. He leads strategy, product, and the company’s scaling trajectory.
CEOCTOGenentech AlumniPrescient Design
N
The Team
Industry Veterans · Life Science AI
NostrumAI’s team is drawn from the frontlines of AI-for-biology — including Genentech’s Prescient Design and leading academic ML groups. Each member brings rare dual expertise: rigorous life science understanding and production-grade machine learning capability.
Drug DiscoveryComputational BiologyML Architecture
// Heritage
The founding team comes directly from Genentech’s Prescient Design — the group that helped establish AI-driven drug discovery as a serious scientific discipline. That heritage is embedded in NostrumAI’s culture: rigorous, experiment-grounded, and scientifically uncompromising.
Our Story

Five years of focused building

2020 — Founding
NostrumAI is established
Founded by alumni of Genentech’s Prescient Design, NostrumAI is built on the thesis that neural scaling laws can be applied to biological problems at the scale of drug discovery.
2021 — Architecture
Foundation models for biology
First generation biological foundation models are trained and validated. The design-run-interpret cycle is prototyped — establishing the closed-loop architecture that defines the NostrumAI platform.
2022–2023 — Platform Build
Autonomous research engine goes live
The autonomous laboratory workflow system is deployed. NostrumAI integrates AI prediction with physical experimental execution — and the loop closes for the first time. Preclinical workflow management is productionised.
2024–2025 — Growth
Revenue scales to $50M
NostrumAI crosses $50 million in annual revenue — a milestone that validates the platform’s commercial depth. Models compound in capability. Cycle times compress. The pipeline expands.
2026 — Now
Established and scaling
NostrumAI is an established AI biotech platform at $50M revenue and growing. Five years of compound learning — in models, data, and operational depth — position the company at the frontier of autonomous drug discovery.