Antonio Osorio

DevOps at Amplitude

This is a talk about the lessons we learned building a healthy DevOps culture, the role of the DevOps team, the tools we lean on, and the challenges ahead. It was a snapshot of the evolution of software development at Amplitude during my tenure as head of cloud engineering.

The Advantages of Real-time Collaborative Design

A project team can generate new ideas far more quickly than it can record and analyze each idea; also, great ideas can happen anytime, not just during regularly scheduled meetings. Every dropped idea is a missed opportunity to find a path forward through the complex maze of drug discovery.

LiveDesign allows every idea to be captured, shared, analyzed, and prioritized—leading to a fuller exploration of the chemical space while facilitating better communication across the different functional groups. Any project team member can enter an idea and instantly get feedback using computational models to help further refine the design—resulting in real-time collaboration across time and location barriers.

My primary responsibility was designing, developing, and maintaining the infrastructure that runs the scientific simulations to inform design decisions.

Achieving highly potent binding—while maintaining a host of other ligand properties required for safety and biological efficacy—is a primary objective of small molecule drug discovery.

Seeing this unmet need, we embarked on a multiyear research project to develop a new free energy calculation technology (FEP+). Our objective was to provide a rigorous approach for computing binding free energies that offers significant value to industrial drug discovery efforts. We are pleased to report that, after utilizing the FEP+ technology on seven different active drug discovery collaborations over the past year, we now have firm evidence that the free energy approach developed in FEP+ can facilitate better synthesis decisions during lead optimization.

My responsibilities were to do the initial, on-cloud proof of concept by automatically scaling FEP+ simulations in AWS while minimizing infrastructure costs.