Vijaya Kaushik
Executive Director, Data Science at JPMorganChase
AI and data product leader with over 20 years of experience, now executive director of data science at JPMorganChase working on enterprise AI, generative AI and analytics across wealth management and stock plan management. She spent seven years as a senior vice president at the bank, earlier led enterprise data management for anti-money-laundering work at Infosys, and was named a Women's Bond Club Rising Star in 2024.
Key arguments & positions
- States plainly that AI is only valuable when it solves a real business problem.
- From her conference talk: multi-agent enterprise platforms should be built evaluation-first, the way disciplined teams write tests first.
- Leads enterprise and generative AI product work inside wealth management and stock plan management.
- Came to AI through enterprise data management and data strategy, including anti-money-laundering work.
Accomplishments
- Executive Director, Data Science at JPMorganChase (2026), after seven years there as a senior vice president.
- Named a Women's Bond Club Rising Star in 2024.
- Led enterprise data management strategy for anti-money-laundering work while at Infosys.
- 20+ years in AI and data product leadership.
Papers & key writings
No published papers or recorded talks found beyond her conference appearances, so her positions above are drawn from her own public role description and talk title.
Links
In the library
Nothing of theirs is filed in the hub yet. Browse the full library.
Recommended next
Hand-picked from the hub based on what Vijaya Kaushik covers.
Real-Time Data Migration: What Worked, What Didn't
Real-time data streaming for AI: a global chip manufacturer shares what worked, what failed, and what they'd do differently. Includes CIO insights on evaluation, implementation, and vendor selection for enterprise AI data infrastructure.
Why this: Also about enterprise AI
Optima — Custom AI Benchmark Builder (Artificial Analysis)
A platform from Artificial Analysis for building custom benchmarks from your own files, agent traces, or coding environment, then running them across leading models to compare quality, cost per task, and time per task. Benchmarks can be graded against objective rubrics or pairwise judging. Optima is a commercial product; the public announcement and product overview are free to read.
Why this: Also about evaluation
abliteration.ai — Unrestricted AI API and Policy Gateway
An OpenAI-compatible API for unrestricted language models aimed at red teaming, security research, evaluations, and synthetic data, paired with a policy gateway for per-project keys, audit logs, and no data retention.
Why this: Also about evaluation
Are You in the Weights? — Model Name-Recognition Experiment
Public experiment testing whether language models appear to recognize a name without web search. Its methodology clusters generated descriptions and combines model confidence into a strength score. The creators warn that confidence is uncalibrated and biographies can be hallucinated; the score is not a verified measure of fame or evidence of a specific training record.
Why this: Also about evaluation