June Meetup

Join us for the June Bay Area OWASP Meetup, proudly sponsored by Semgrep.
Spend the evening connecting with fellow security professionals, learning from insightful presentations, and engaging in meaningful discussions on the latest security topics. Enjoy great networking opportunities along with food and refreshments, generously provided by Semgrep.

5:00 PM : Doors open, Networking, food and drinks
5.30 PM :- Chapter introduction
5.45-6.30 PM :- A Strategy for Evaluating LLM + SAST
6.30-7.15 PM :- Frontier Labs Said It Was Build, Fixed and Implemented! What can go wrong.
7.15-8.00 PM :- The Security Tab You're Running Up: What AI-Generated Code Costs in the Long Run

Talk#1: A Strategy for Evaluating LLM + SAST
Abstract: LLMs can reason about security bugs, but when pointed straight at a codebase they're inconsistent, expensive, and noisy. The biggest gains come from what wraps the model. We show how the right harness can drive recall, precision, and run-to-run consistency. We benchmark this with models like Claude Opus 4.8, GPT 5.5, etc. to determine how best to drive down the price per true positive vulnerability.

Speakers:- Shelley Wu, Senior Product Manager and Daghan Atlas, Head of Product

Talk#2: Frontier Labs Said It Was Build, Fixed and Implemented! What can go wrong.
Abstract: As AI becomes part of software delivery, teams increasingly rely on vibe coding for security recommendations. This talk examines why confidence is not evidence and presents approaches for independently validating AI-driven security decisions.

• AI trust boundaries
• Verification of AI outputs
• Security agent governance
• Evidence-backed automation

Speaker Bio: Swapnil Deshmukh is a cybersecurity executive, entrepreneur, and security practitioner with more than 15 years of experience leading security, application security, cloud security, and product security initiatives across financial services, payments, and technology organizations. He has held security leadership roles at a few Fortune 500 companies and has worked on securing large-scale platforms, software delivery systems, and enterprise infrastructure.
Swapnil is a co-founder of Certus Cybersecurity and has advised organizations ranging from startups to Fortune 500 companies on application security, cloud security, identity, and secure software delivery. His interests include modern software engineering, AI-assisted development, security governance, and building scalable approaches to security decision-making.

Talk #3 "The Security Tab You're Running Up: What AI-Generated Code Costs in the Long Run"

Abstract: AI coding tools have made it easier than ever to ship fast, but the security bill arrives later and lands far from where the velocity wins were celebrated. In this talk,

Speaker: Ankit Agrawal, Sr. Security Engineering Manager at Webflow, draws from real-world experience to map out where the hidden costs actually concentrate across engineering organizations, independent developers, and marketplace ecosystems.
The session covers the growing asymmetry between how fast code is being generated and how slowly security practices are keeping up, what defender burnout looks like at scale, and the second-order risks that are easy to miss until something breaks. It closes with a prioritized, practical framework for reducing the cleanup cost based on patterns that are actually working in production environments.