It’s no secret that artificial intelligence is transforming the way we work, learn, and create. But not every revolution starts at the top. To mark International AI Day (July 16), we decided to share three stories from Israeli tech companies that shed a different light on AI adoption: employees who spotted an opportunity, developed an idea, and launched AI-driven projects that became catalysts for change both within their organizations and beyond.
These stories illustrate how the democratization of AI tools opens the door to decentralized initiatives, born out of curiosity and a sense of purpose, and provide organizations with a new infrastructure for learning, automation, and social impact.
SAP R&D Center: An AI guild born from the ground up, reaching Stanford
At SAP’s R&D Center in Israel, an original grassroots initiative was born: an internal “AI Guild” launched by developers seeking to deepen their expertise in this domain. What began as a learning need quickly evolved into a dynamic, cross-functional community.
Asaf Bruner, SAP (Photo: Courtesy)
“I co-founded the guild with Daniel Nemeth, Yosi Izaq, and Dina Vinter, all senior developers here at the center,” says Asaf Bruner, Engineering Group Manager at SAP. “Early on in the GenAI wave – and particularly in the Business AI space – we sensed that something big was happening. It wasn’t just about keeping up and learning more, it was about building an intra-organizational community that would help us stay ahead in a field that’s evolving at breakneck speed.”
The guild (named after the traditional professional collectives) was founded on the belief that the most effective way to learn in a fast-moving space like GenAI is together, through knowledge-sharing, trial-and-error, and multidisciplinary dialogue.
“We started small. A few people who were into it and wanted to delve deeper. Today, we’re more than 40 members, and we constantly get new requests to join, but we intentionally keep the group relatively small so that everyone remains an active contributor. This isn’t a listening club; it’s a space for hands-on work, real contributions, and meaningful initiatives,” Bruner says.
The guild is open to all employees at SAP’s R&D Center in Israel, starting from developers, through product managers, and up to team leads. “The goal is to look at the entire development chain through an AI lens. From idea to deployment to customer impact,” Bruner explains. “It’s not just a group of ‘AI geeks,’ but a mix of individuals from all corners of the organization, with diverse professional backgrounds. That makes the conversations much richer and the insights much deeper.”
Fortunately for them, the initiative won immediate backing from leadership. “From day one, we had strong support from the center’s management, especially Managing Director Orna Kleinmann, who embraced the initiative even when it was just an idea. She gave us resources, organizational connections, and a great deal of inspiration.”
The guild already managed to initiate cross-team sessions, fostered collaborations with foreign and domestic units, and even sent representatives to Silicon Valley, where they visited Stanford University and met with leading AI researchers. “We met trailblazers like Prof. Kanishk Gandhi, who explores inference capabilities in AI models, and Prof. Igor Denisov, who studies developer productivity and the impact of AI tools,” says Bruner. “It was fascinating and opened our minds to how AI can truly be measured, optimized, and put into practice.”
But above all, for Bruner, the most significant achievement is the cultural shift: “In a relatively short time, we managed to change the organizational culture. Introducing real innovation, connecting people, and giving the organization practical tools to navigate a technological revolution that is happening right here. Right now.”
Commit: AI born in crisis, built for routine
At the tech company Commit, an AI initiative also emerged “from the ground up,” this time out of crisis. When the “Iron Swords” war broke out, about half of the Company’s employees were called up for reserve duty, making the need for fast, smart, and automated solutions to critical business tasks absolutely critical.
Ohad Dallal, Commit (Photo: Nati Levi)
“We were facing a complex operational crisis. Many employees were called up for reserve duty under emergency orders, with no way to predict when they would return,” says Ohad Dallal, co-founder of Commit and currently head of Business AI Operations at the Company. “The immediate technical challenge centered on our customer support system: how to identify, in real time, clients who urgently needed assistance, even if they had not flagged the ticket accordingly, without overburdening the remaining team with extra reporting and tracking. Fortunately, our automation team jumped in right away to develop solutions that would ensure service continuity by deploying automated tools to compensate for the absence of key personnel.”
The system they built relied on AI Agents capable of detecting urgent tickets based on free-text inputs and triggering automatic alerts. Over time, these agents became a core element in the Company’s overall operations.
“Today, the system is implemented across the organization in a decentralized manner, based on functions: business unit managers receive agents tailored to their unit’s data, while team leads, customer managers, and project managers get agents adapted to their specific content domain.”
Consequently, the Company now defines an “Organizational Agent Hierarchy” and adopts a hybrid operating model composed of both employees and agents, fully embracing the technological paradigm at the organizational level.
According to Dallal, the impact of this move is still felt across all operational levels: “The fundamental shift is the dramatic reduction in unnecessary internal communication and the creation of a fast, unified interface for organizational data. Instead of addressing specific employees through Teams (for example, the accountant) with inquiries pertaining to invoice status or credit balance, the system enables interaction with a virtual agent via chat. The agent gathers and displays the required information automatically and instantly. The agents are connected to all company systems and allow fast, simultaneous access to multiple platforms.”
“By the way, the first solution we developed – a ticketing-analyst agent who identifies emergencies among clients [detecting phrases like ‘I need you to call me back urgently’] is still in use, sending automatic alerts to the support manager, the team, and relevant stakeholders based on context,” Dallal concludes.
AI21: Back to school 2.0
The initiative led by Shuki Cohen, VP of Data at unicorn startup AI21, began in the most unexpected place: his son’s elementary school. Before long, it evolved into a small educational community centered around AI.
“Whenever I mention that I work at a company developing artificial intelligence, people’s eyes and ears immediately perk up,” he says. “One morning at my son’s school in Jerusalem, his teacher got excited about what I do and suggested I give a talk to the students. I jumped at the opportunity, knowing it would be an empowering moment for my son.”
Shuki Cohen, AI21 (Photo: Roy Shor)
Cohen made sure to make the content of his talk accessible for children, aiming to keep the lesson “light and fun, and much less technical or academic.” The students’ responses were positive, and even surprising. “The kids were really excited to learn, and to be honest, they impressed me with how much they already knew,” Cohen says. “One student even raised an interesting point about copyright in the context of AI – not something you’d expect from a second grader.”
After the classroom experience, Cohen shared the lesson plan in a LinkedIn post, and the responses flooded in. “I didn’t expect it to get so much attention,” he says. “People from very different backgrounds – not just those in AI – asked for the content. Everyone showed genuine interest. I think the growing public conversation around AI creates a real thirst for knowledge, and people are eager to pass it on to students. It’s the kind of thing that feels like hitting two birds with one stone.”
The strong response sparked in Cohen a desire to expand his efforts in the field. “There’s clearly a major gap in how we teach and expose children to AI,” he says. “Which is why I have already started thinking about additional content I’d like to share with the growing community – including a glossary of terms, short explanations, data privacy considerations, and more.”
He also has broader, systemic ambitions: “I’d love to contribute to AI education on a national level, in collaboration with the Ministry of Education. Right now, it’s just an idea, but it’s clear we are in the midst of an AI storm, and the younger generation isn’t necessarily prepared for these radical changes.”
