Your Evals Passed. Your Agent Just Emptied a Database.
- Discuss this with your agent
- Open in Claude
- Open in ChatGPT
Offline evaluations won't stop a deployed AI agent from hallucinating tool calls and corrupting production data. Learn how runtime verification layers can keep your databases safe.
Checking access…
Playback and chapters load privately for Free videos.
Matching moments
More from World Congress 2026 North America
Upcoming sessions on this topic
Starts in 5 days
Conference India 2026
November 25–26, 2026
Bengaluru, India
-
Autopsy of an Autonomous Incident: When the Agent Made It Worse
Navin Pai · StackGen
Wed 25 Nov · Stage 1 - powered by IBM
-
The Developer’s Guide to the Agentic Git Forge: Agents That Remember
Haimantika Mitra · Entire
Wed 25 Nov · Room 7
-
Your AI Agents Are Bleeding Tokens: Building a Runtime Cost Control Plane for LLMs

Aakriti Aggarwal & Bhavishya Pandit
Wed 25 Nov · Stage 1 - powered by IBM
Related videos
Related articles
DC
Daniel Cranney
DC
Daniel Cranney
IK
Igor Khokhriakov
CH
Chris Heilmann
TP
Thomas Pamminger
From learning to earning
Jobs that call for the skills explored in this talk.
5 days ago
Staff Software Engineer, Copilot Code Review
GitHub
Beverly Hills, CA, United States
Experienced
Remote
Ruby
Golang
Github
2 months ago
•
Principal Product Manager, Agentic Evals
Expedia
San Jose, United States
Expert
Hybrid · 3 days on-site
Product Management
10 days ago
Principal Software Engineer, Copilot Code Review
GitHub
Beverly Hills, CA, United States
Expert
Remote
Ruby
Golang
Github
18 days ago
•
Senior Software Engineer - Semantic Data Lake (Open)SDE3
WEX
Bengaluru, India
Expert
Hybrid · 4 days on-site
Scala
Python
Terraform
26 days ago
Principal Software Engineer, Copilot Code Review
GitHub
San Francisco, CA, United States
Expert
Ruby
Golang
Github
2 months ago
•
Principal AI Forward Deployed Engineer
Expedia
San Jose, United States
Expert
Hybrid · 3 days on-site
AI Frameworks
Software Architecture