Conference affiliation: Engineering Group Manager · monday.com · 2026
On this page
Omri Bruchim is an engineering leader at monday.com building monday sidekick, an AI assistant that helps people understand and prioritize their work. His career spans mobile infrastructure, no-code software, real-time data systems, and enterprise AI.
From mobile platforms to enterprise AI
Bruchim studied at Ben-Gurion University of the Negev and earned an MBA from Tel Aviv University. After engineering and mobile leadership roles at Soluto and Asurion, he joined Wix, where he worked on mobile infrastructure, React Native, and application performance. His open-source react-native-perf-logger reflects that early focus on understanding production software.
He later co-led Branded App by Wix, which enables businesses to create native mobile applications without writing code. Building the product involved organizing engineering, design, analytics, and product teams around direct relationships with customers, as described in Wix’s account of its mobile app builder.
Bruchim subsequently co-founded drift.dev, serving as chief technology officer of a startup focused on monitoring and controlling real-time data pipelines. He brought that experience with streaming infrastructure and rapidly changing information into his AI work at monday.com.
Teaching assistants what work actually means
Bruchim’s central argument is that connecting an assistant to workplace applications does not teach it which commitments matter or how projects, people, and decisions relate. His work on Sidekick concentrates on three practical ideas:
Systems of context:Enterprise assistants need relationships between records, not simply more search results. A code change, for example, becomes meaningful when connected to its pull request, project task, and the customer problem behind it.
Monday world model: A contextual representation of people, projects, dependencies, and commitments gives Sidekick a working understanding of an individual’s responsibilities. Bruchim describes that effort in his account of building the world model.
Precomputed work context: Bruchim and colleague Tomer Ast describe an architecture that combines durable behavioral patterns with fast-changing signals, isolates unreliable sources, and checks important information against live data. Their AI Engineer conference talk also acknowledges cold starts, biased signals, and the difficulty of identifying genuinely consequential developments.
Bruchim’s interest also extends to how an agents-first world affects people, connecting production AI design with the human consequences of increasingly capable assistants.
