She launched her career at Google Research/Google Brain, contributing to sequence transduction models and the Transformer architecture, foundational work that propelled major advances in NLP and multimodal AI.

Her formal path includes a B.E. from Pune Institute of Computer Technology and an MS in Computer Science from USC, with early self-directed study guided by leaders like Andrew Ng and Peter Norvig.

In late 2024 she joined Anthropic as technical staff (publicly announced in early 2025), focusing on reinforcement learning, frontier model capabilities, test-time scaling, and interpretability, contributing to Claude 3.7 Sonnet.

As an Indian-born software engineer, she co-authored the influential Transformer paper and now shapes Claude and its reinforcement learning capabilities at Anthropic.

Her story underscores how Indian talent has become central to the global AI revolution, blending deep technical contributions with entrepreneurial leadership at the AI frontier.

She extended self-attention concepts beyond language to vision and speech, contributing to Image Transformer, Bottleneck Transformers, and Conformer models, illustrating cross-domain impact.

After Google, she co-founded Adept AI Labs in 2022 as CTO to develop action models for automating enterprise software, and Essential AI in 2023 to build full-stack AI products for data-heavy business tasks.

Her Google Scholar citations exceed 315,000, signaling broad influence on the ML community and foundational impact on models used by OpenAI, Google DeepMind, and academia.