AI agents are meant to accelerate our work—not hold us back. Yet for senior practitioners working daily with agents and coding companions, the shift has been profound: we are moving from makers to managers. This shift introduces a hidden Management Tax—the cognitive load of coordinating outputs, curating context, reviewing multi-step reasoning, and defining trust boundaries that can quietly erode the productivity AI promises.
This round table is designed for power users and team leads navigating the messy middle of AI implementation. Moving beyond prompt techniques, we will focus on shifting from reactive oversight to intentional orchestration of agent-driven workflows.
Together, we will examine:
The Audit Paradox — How to evaluate deep, multi-step agentic outputs without recreating the work manually.
Context Silos & Drift — How to manage context across multiple agents without duplication or hallucination.
Minimal Viable Management (MVM) — How to define high-stakes vs. low-stakes protocols to determine when deep auditing is required and when “trust but verify” is sufficient.
Participants will identify where their own Management Tax accumulates and explore structural approaches for reducing overhead without sacrificing rigor.
The goal is not more AI usage, but sustainable leverage—ensuring AI remains an amplifier of expertise rather than a source of hidden burnout
Alice is a Data Science Team Leader with over eight years of experience building and leading advanced analytics initiatives in media and ad tech. She holds a Master’s degree in Applied Statistics and combines strong theoretical foundations with hands-on technical leadership.Recognized for innovation and impact, she was named Top Women in Media & Ad Tech (Data Demystifiers category). Alice is Co-COO of the WiDS (Women in Data Science) community, where she helps drive initiatives that promote inclusion, professional growth, and knowledge sharing across the data ecosystem.An experienced public speaker, she has delivered industry talks including “A Brief History of Data Science,” translating complex technical ideas into accessible, practical insights. A returning WiDS TLV roundtable speaker, she continues to contribute to the evolving conversation on the future of data science and AI leadership. Alice is passionate about fostering collaboration and helping teams navigate technological shifts with clarity, rigor, and long-term thinking.
Hagit is a Senior Data Scientist and veteran machine learning researcher specializing in ranking systems and preference modeling. With deep expertise in reinforcement learning and large-scale optimization, she designs high-impact ML solutions for complex, high-throughput environments, including ad optimization at Arpeely.
Her experience spans diverse domains, from Kubernetes cost optimization and log-based root cause analysis at Microsoft, to professional ranking at Fiverr and cyber risk scoring solutions. Throughout her career — including research collaborations during her PhD with IBM — she has translated complex business challenges into end-to-end machine learning systems, leading initiatives from ideation to production deployment. An experienced conference speaker, she regularly shares insights on ranking, optimization, and applied AI, bridging rigorous research with real-world impact.
Keynote session: Hadas Grossmon Ella
Break
Lightning talks session
Roundtable closing
Talk by Hila Paz
Talk by Dr. Moran Mizrahi
Closing remarks
End
