Research computing
Technology and research-computing leadership
Science and data only move as fast as the organizations that support them. Gabriele Fariello builds and turns around the technology and research-computing organizations that make that possible, inside high-scrutiny institutions where the technical and governance difficulty is the hard part.
The problem: fractured or greenfield technical organizations
Research-intensive institutions often depend on technology and research-computing organizations that are fractured, under strain, or not yet built at all. Costs rise while capability stalls, relationships between central IT and the faculty or scientists it serves grow strained, and the computing that research depends on falls behind the demand. In a high-scrutiny setting the leader has to fix all of that without breaking the work that is already running, and while winning the trust of an expert, demanding audience.
How Gabriele approaches it
Gabriele reorganizes the operating model rather than defending an expensive, siloed status quo. He expands what the organization can do, restores confidence in it, and lets the economics improve as a consequence. He wins by hiring, turning around, retaining, and motivating exceptional people, and he is technically deep enough to hold his own with the scientists and engineers he leads. The result is capability expansion and trust restoration together, produced by disciplined operating leadership under pressure.
The proof
At the Harvard School of Engineering and Applied Sciences, Gabriele was the inaugural CIO and Assistant Dean for Computing, asked to rebuild a fractured IT and computing organization, mend a strained relationship with the wider faculty over research computing, and bring rising costs under control. He cut real IT spend by nearly 30% over three years while services expanded, delivered up to roughly $13.9M in cumulative savings and avoided cost, and drove per-capita IT spend down 31 to 44 percent at a time when no peer Harvard organization achieved reductions greater than 10 percent. In the same period he expanded the research computing available to faculty and students roughly fifty-fold at effectively no added cost, by consolidating the school's high-performance computing into Harvard's shared research-computing environment.
He also took the school fully to the cloud, the first at Harvard to do so, and announced it at the University CIO Council two weeks after a major 2013 AWS outage, when the move was still contested. The reasoning was about accountability and economics. The school's data centers ran below 99% uptime and left it fully liable when they failed, while a hyperscale provider offered reliability and durability guarantees the school could not match alone. He worked with Harvard's University Chief Information Security Officer to develop the cloud-security standards that other schools later adopted. About 85% of critical infrastructure moved to the cloud, estimated uptime rose from roughly 98.6% to about 99.8%, and the real payoff was leverage: he redirected roughly three and a half full-time staff from commodity infrastructure to research and academic computing, reaching total-cost parity with far more value delivered for the same spend.
Earlier, Gabriele co-founded and led Harvard's neuroinformatics group, building it into an internationally recognized effort and designing the data infrastructure and automation that let neuroscientists process brain-imaging data at scale, with the reproducibility rigorous science demands. He supported the informatics behind large-scale, funded neuroscience, including work tied to a human-connectome research grant on the order of $30M. He won the confidence of a faculty that was deeply skeptical at the outset, in part because he did not hold a terminal degree, to the point that they voted to fund the position from their own departmental resources. A skeptical research faculty does not extend that kind of endorsement lightly.
Gabriele held that neuroinformatics role and a simultaneous senior appointment in clinical research informatics at Massachusetts General Hospital at the same time, serving a large community of clinicians and researchers. Carrying both is itself a proof of trust and capacity.
Read the two case studies that carry this focus: the Harvard SEAS computing turnaround and Harvard neuroinformatics.
Why it transfers
Building or turning around a technical organization in a high-scrutiny institution is the same capability that a regulated-industry CAIO, CIO, or AI-transformation mandate requires: expand what the organization can do with data and computing, win the trust of skeptical technical and scientific stakeholders, and hold financial discipline at the same time. Gabriele has shown all three at institutional scale, and the scale shows through the complexity of consolidating and building research computing rather than through raw headcount.
Harvard SEAS case study · Harvard neuroinformatics case study