Speaker: Chris Perkins, Director of Lab IT and Operations, Broad Institute

Zifo’s SiEE event 2026

The Ultimate Mission: Navigating Lab IT in a Highly Heterogeneous Academic Environment

In a presentation on operationalizing data flow, Chris Perkins, Director of Lab IT and Operations at the Broad Institute, highlighted the monumental task of managing IT assets and data pipelines in a decentralized, academic environment. Transitioning from industry to academia, Chris discovered that instead of top-down corporate mandates, the Broad operates more like a collection of “fiefdoms” where researchers dictate their own funding, hardware, and networks.

The ultimate goal for his team is to move away from messy, disconnected data practices toward highly structured, interoperable datasets — the necessary foundation for AI models like AlphaFold. In this presentation, Chris explores the delicate balance of building a metadata-first culture, driving standardization without authority, and creating practical lab interventions that prioritize the researchers’ immediate scientific needs.

Key Takeaways from the Presentation

• Building a “Metadatism” Culture: You cannot simply shove data into an LLM and expect revolutionary results. Cultivating a culture-like “metadatism” means capturing metadata at every step using structured schemas. Critically, systems must be designed so scientists never have to enter metadata more than once, minimizing the administrative burden at the bench.
• The “Sneakernet” Reality and Tactical Solutions: Because scientists often face heavy security friction when trying to move data to the cloud, they frequently rely on USB drives or buying private servers. Rather than fighting this directly, Chris’s team meets users where they are. They implement “mule architectures” (jump boxes) to safely extract data from older lab computers and use automated background scripts that mimic the simplicity of a USB transfer, quietly routing data directly to the cloud.
• Financial Transparency and Cloud Trust: To address researchers’ fear of unpredictable cloud fees, Chris uses his business background to present a clear cost comparison over three years. He walks them through the true cost of buying and maintaining a local NAS drive — including the reality of having to purchase another one once it fills up — versus a Cloud subscription. This transparency is part of an ongoing effort to rebuild trust in cloud storage after past instances of unexpected bills left scientists wary of the technology.
• The Carrot Over the Stick for Standardization: In an environment with zero central ELN or LIMS mandates, IT cannot simply force compliance. To drive the adoption of uniform data platforms, IT must build exceptional workflows that generate a “fear of missing out” (FOMO). By automating data ingestion directly at the instrument, IT removes the manual data entry barrier, encouraging natural adoption.
• Balancing Security and Science: IT teams are sometimes viewed as hostile invaders who get in the way of science. Chris emphasizes that the IT mindset must pivot from immediate control to collaborative support. By physically sitting side-by-side with scientists in the lab to solve problems, IT builds necessary trust, ultimately proving that good security and good science can coexist.