DJ Kleinbaum

Company and Platform Overview

• Emerald Therapeutics, founded five years ago by two PhDs in life sciences and computer science from Carnegie Mellon University, pivoted from an antiviral therapeutic focus to the Emerald Cloud Laboratory (ECL).
• The ECL is a fully virtualized life sciences facility located in South San Francisco that allows scientists globally to design and run experiments remotely.
• The platform distinguishes itself from traditional Contract Research Organizations (CROs) by focusing on virtualization rather than outsourcing, aiming to provide users full control over experimental parameters rather than bundling consulting services.

Operational Model and Efficiency

• The company raised a significantly smaller initial capital than the typical $20–$40 million Series A for biotech startups, necessitating a radical approach to maximize scientist efficiency through automation.
• Every experiment at the facility is fully automated ("push a button, walk away"), eliminating the need for scientists to babysit instruments during runs.
• Scientists are trained to write code in the Wolfram language to build tools and solve specific research problems, bypassing low-level programming concerns.
• The system treats experimental protocols as data objects within a unified network, effectively automating sample tracking, instrumentation diagnostics, and inventory management.

Data Architecture and Reproducibility

• The ECL captures and stores all ancillary and meta-data for every experiment, including instrument maintenance logs, control run histories, and real-time environmental variables (e.g., room humidity and temperature).
• The platform utilizes a strict ontology to standardize data fields across different instruments, ensuring consistent naming conventions (e.g., "side scatter") regardless of the hardware used.
• By recording every scientifically relevant variable robotically, the system aims to solve the scientific reproducibility crisis, enabling "push-button reproducibility" of results.
• Data is structured as a dense, scale-free network where data objects are linked to the specific experiments, instruments, and sample histories that generated them.

User Interface and Workflow

• The design interface features a dual-mode system: a graphical user interface (GUI) for point-and-click experimentation and a machine-readable command line that functions as "hotkeys" for power users.
• The system employs an "intelligent defaults" engine where Scientific Development teams configure optimal parameters based on selected reagents, allowing users to specify only the "diffs" (deltas) for their specific needs.
• Once a protocol is submitted, it enters a queue with a target 48-hour turnaround time to facilitate rapid research and development iteration.
• Returned data includes not only experimental results but also a complete lineage of the sample, previous related experiments, and instrument health metrics.

Current Status and Strategic Outlook

• The ECL transitioned from an internal efficiency tool to an external service over the last 15 months to monetize the platform's value beyond the company's original research goals.
• Public beta access began at the very end of Q1, with the company currently ramping up capacity to onboard more external scientists.
• The platform includes a suite of analysis tools designed to handle the volume of data generated by high-throughput, automated experimentation, moving beyond standard spreadsheet-based analysis.