Prashant Kulkarni

CEO

San Francisco, California, United States26 yrs experience

Most Likely To SwitchAI ML Practitioner

Key Highlights

Expert in AI security and large language models.
Extensive experience in Google Cloud security architecture.
Passionate about innovation and risk mitigation.

Stackforce AI infers this person is a Cloud Security Architect specializing in AI safety and compliance for enterprise solutions.

Contact

Skills

Core Skills

Trustworthy MlAi/ml SecurityGoogle Cloud SecurityArtificial Intelligence (ai) Safety & TrustHigh Performance Computing (hpc)TensorflowData PrivacyKubernetesDockerGenerative AiMultimodal Models

Other Skills

University LecturingAI/ML/Data workloadssecurity researchlarge language modelssecurity riskscustomer collaborationKerasResearch SkillsNatural Language Processing (NLP)Large Language Models (LLM)Responsible AIPython (Programming Language)Vertex aiGoogle Cloud Platform (GCP)Inpsect

About

As Lead AI Security Research Engineer specializing in LLM security at Google, I blend hands-on cybersecurity expertise with advanced academic research to safeguard next-generation AI systems. I bring extensive experience in various aspects of Google Cloud security and security architecture. My current research focuses on identifying and mitigating vulnerabilities in large language models, ensuring the integrity and resilience of AI-driven technologies. Passionate about innovation and risk mitigation, I’m dedicated to developing comprehensive defenses in today’s dynamic threat landscape.

Experience

26 yrs

Total Experience

2 yrs 8 mos

Average Tenure

6 yrs 7 mos

Current Experience

Constellation research center

Astra Research Fellow (AI Safety) / Redwood Research

Jan 2026 – Mar 2026 · 2 mos · Berkeley, California, United States · On-site

On sabbatical from Google to conduct AI safety research under Redwood Research mentorship as part of the Astra Fellowship at Constellation's Berkeley research center.

Supervised program for alignment research

Spar AI Research Fellow

Sep 2025 – Dec 2025 · 3 mos · Remote

Contributing to SandboxBench, a security evaluation benchmark built on UK AISI's Inspect that tests whether AI agents can escape their containment boundaries. As AI systems gain autonomy, understanding sandbox escape risks becomes essential for safe deployment.
This work directly supports UK AISI's mission to evaluate frontier AI capabilities and risks at scale. The benchmark now enables high-volume batch evaluations across 27 challenge scenarios spanning container escape, data exfiltration, and lateral movement.
https://github.com/UKGovernmentBEIS/inspect_evals/tree/main/src/inspect_evals/sandboxbench

InpsectKubernetesDocker

University of chicago

Teaching Assistant

Sep 2025 – Dec 2025 · 3 mos · Remote

Teaching Assistant for Dr. Fouad Bousetoune for Applied Generative AI : Agents and Multimodal Models for Fall 2025

AgenticGenerative AIMultimodal models

Ucla extension

Adjunct Instructor

Jun 2025 – Present · 11 mos

I teach a series of courses in Trustworthy ML and Generative AI
Trustworthy ML - https://trustworthyml-ai.github.io/

University LecturingTrustworthy MLAI/ML SecurityGenerative AI

Google cloud security

Lead AI Security Research Engineer

Mar 2025 – Present · 1 yr 2 mos · Sunnyvale, California, United States

I primarily work on various aspects of Google Cloud security for AI/ML/Data workloads for Google's enterprise customers.
I split my time with security research focused on large language models (LLMs) and the vulnerabilities they face. My responsibilities involve not only conducting in-depth analyses to identify and mitigate security risks but also collaborating directly with our customers to tailor solutions that address their unique challenges. This role allows me to bridge the gap between advanced research and practical application, ensuring that our AI systems are both innovative and secure.

Google Cloud securityAI/ML/Data workloadssecurity researchlarge language modelssecurity riskscustomer collaboration+1

Google

3 roles

Cloud Security Architect

Aug 2024 – Present · 1 yr 9 mos

Artificial Intelligence (AI) Safety & TrustKerasResearch SkillsGenerative AINatural Language Processing (NLP)Large Language Models (LLM)+3

Google Fellow at Urban Systems Lab

Feb 2024 – Aug 2024 · 6 mos

Working with Urban Systems Lab to develop climate modeling for ClimateIQ.
First part of this assignment I’ve worked on high performance computing (HPC) to productionize Weather Research Forecasting (WRF) physics model on Google Cloud for label generation.
Second part of this assignment I’m working on developing a tensorflow training pipeline to train flood modeling on Vertex AI.

Python (Programming Language)Vertex aiHigh Performance Computing (HPC)Google Cloud Platform (GCP)TensorFlow

Google Cloud Security & Compliance Specialist

Oct 2019 – Aug 2024 · 4 yrs 10 mos

Security and Compliance specialist at Google Cloud Sales Engineering.
Help Google Cloud customers secure their workloads on Google Cloud - infrastructure, data, IAM SecOps and AI/ML

Data Privacy

Ripple

Security Architect

Sep 2018 – Sep 2019 · 1 yr · San Francisco Bay Area

AWS & Google Cloud IAM, Cloud Security, App sec and container security.

Data Privacy

Bank of the west

Security Solutions Architect

Jan 2016 – Sep 2018 · 2 yrs 8 mos · San Francisco Bay Area

Qualcomm

Sr. Staff Engineer/Manager, IT

Oct 2012 – Jan 2016 · 3 yrs 3 mos · San Diego County, California, United States

Hired and managed IAM engineering team in San Diego and Offshore locations

Oracle

2 roles

Consulting Technical Manager

Sep 2011 – Oct 2012 · 1 yr 1 mo

Worked as a lead solutions architect for Oracle Consulting Services. Handled implementation of Oracle Identity Management suite for various customers of Oracle.