Infant Rosario Vasanth is a technical fellow and senior director of engineering at BlackRock, leading work on the Studio Compute Platform and Data & AI Acceleration. He builds the infrastructure that turns financial research, prospectuses, and other complex investment documents into usable analysis and automated operational workflows.
His earlier career included roles at Kable, Squrl, IndustryNext, and iSpeech. He studied computer science and engineering at Birla Institute of Technology and Science, Pilani, and earned master’s degrees from Heriot-Watt University and the New Jersey Institute of Technology. He joined BlackRock in 2011 and eventually became a technical fellow, with responsibilities spanning research platforms, automation, and Aladdin Studio’s AI capabilities.
With fellow BlackRock engineer Vaibhav Page, Vasanth co-invented a patented Kubernetes-native application deployment engine, granted in 2026. The system uses custom resources, state machines, and controller reconciliation to manage application workloads.
How he approaches financial AI
- Put financial specialists inside the development loop. His sandbox-and-app-factory architecture gives domain experts tools to refine extraction templates, prompts, and model strategies; a cloud-native deployment layer turns those configurations into operational applications.
- Engineer for the document, not the demo. Prospectuses and term sheets vary substantially in complexity and length. Vasanth combines prompt evaluation, configurable extraction, retrieval-augmented generation, and multistep reasoning according to each task, then connects structured outputs to downstream investment systems.
- Match infrastructure to economics. Processing hundreds of research reports may warrant GPU-based inference; individual security-setup workflows can use burstable compute. He evaluates infrastructure, access controls, deployment costs, and whether existing products offer better value than custom applications.
- Keep expert judgment operational. His approach to human-in-the-loop financial AI incorporates review, compliance controls, and security safeguards from the beginning because autonomous agents cannot reliably substitute for specialized knowledge in regulated investment operations. See Vasanth and Page explain the architecture.
