Dor Cohen
AI Engineering Director @ monday.com
Tel Aviv, Israel
Actions
Building the future of work where humans & AI agents collaborate seamlessly.
Leading AI Infra & Agent Ecosystem @mondaydotcom. 🤖🚀
Engineering leader with 20 years of experience in establishing and managing teams of 50 people across all stages of the product lifecycle, with a proven track record in global companies and startups.
Led multiple platform groups at monday.com, responsible for both user experience and infrastructure, serving millions of users and ensuring a scalable and efficient data infrastructure.
Successfully delivered & sold a multi-million-dollar Big Data platform from scratch to enterprise customers in the public Homeland Security and private B2B/B2G sectors. Led Check Point’s Container Security solution, providing a multi-layer cloud workload protection platform (CWPP).
Mamram graduate with 7 years in an elite Intelligence Corps technological unit.
Holds a BA in Computer Science and an MBA in Technology, Innovation & Entrepreneurship (MoTIE) from Recanati, Tel Aviv University.
Area of Expertise
Topics
Closing the Loop: Engineering Reliable AI Agents in Production
As AI agents move from assisting users to executing real work, reliability becomes one of the hardest engineering challenges. An agent can produce a convincing result while missing critical data, skipping actions, or failing to complete the task. Traditional LLM evaluations often fail to detect these problems.
At monday.com, we're building AI agents that work alongside humans across our workspace platform. To make these agents reliable, we developed a continuous evaluation architecture that connects pre-production testing with real production behavior.
In this session, we'll dive into how we combine offline evaluations running against real staging dependencies with online evaluations analyzing production traces. We'll share how we measure task completion, evaluate agent execution paths, detect regressions, and turn production failures into new test cases, creating a continuous improvement loop.
We'll also explore how the same evaluation infrastructure supports benchmarking frontier and open models, optimizing quality, latency, and cost, and our next steps toward fine-tuning specialized evaluation models.
Through real production examples, architectural decisions, and lessons learned at scale, you'll discover practical approaches to building, measuring, and continuously improving AI agents that users can trust.
Key Takeaways:
* Measure what matters for AI agents: Go beyond response quality to evaluate actual task completion, execution trajectories, and reliability in real-world workflows.
* Close the loop between testing and production: Combine offline and online evaluations to catch regressions before release, discover unexpected production failures, and turn them into automated tests.
* Use evaluations to drive model decisions: Benchmark frontier and open models, optimize quality, latency, and cost, and explore fine-tuning specialized evaluation models using real production data.
Close the loop session on LangChain Interrupt London 26' - https://www.youtube.com/watch?v=vd7zSbA_heE
When Agents Become Teammates: Rearchitecting Identity and Permissions for Enterprise AI
What happens when an AI agent is no longer just your assistant, but a teammate working across your organization? And what if that agent was built by someone else?
Most platforms treat AI agents as applications acting on behalf of a user, inheriting their identity and permissions. This works for simple use cases, but breaks down when agents operate autonomously, collaborate across teams, and connect from external platforms. Who owns the agent? What can it access? And who is accountable when something goes wrong?
At monday.com, where we're building a future in which humans and AI agents work side by side, these challenges pushed us to rethink our identity and permissions architecture.
In this session, we'll share how we're evolving our platform to treat AI agents as first-class identities, with their own permissions, access boundaries, and audit trails. We'll dive into the architectural challenges of supporting both native monday.com agents and external agents built on third-party platforms, while maintaining enterprise-grade security and governance.
Through real engineering challenges, architectural decisions, and lessons learned, we'll explore what it takes to move beyond user impersonation and build the foundation for trusted human-agent collaboration at enterprise scale.
Key Takeaways:
* Why traditional permissions break: How inherited user permissions create security risks, unpredictable access, and accountability gaps as agents become autonomous.
* Agents as first-class identities: Designing dedicated agent identities, scoped permissions, provisioning, and auditability.
* Bring Your Own Agent/Key (BYOA/BYOK): How to securely integrate external and customer-built agents into enterprise workflows without compromising control.
* Engineering for enterprise scale: Architectural patterns and lessons for building secure, governed, and scalable human-agent collaboration.
Session presented on WeAreDevelopers, North America 26' - https://www.wearedevelopers.com/world-congress-north-america/speakers/dor-cohen
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top