Startup brief
Hyperscale Demand and Agentic AI Drive Israeli Startup Capital
From construction AI and event databases to AI cybersecurity, Israeli founders attract fresh venture backing across physical infrastructure and agent governance.
Startup
OpenAI paid $300 million for Glass Imaging to anchor its hardware ambitions in custom computational optics — taking direct control over image input quality for multimodal models.
BackgroundGlass Imaging develops neural network algorithms and tailored optics designs that correct physical lens distortions and elevate mobile camera quality. Attar previously led defense R&D at Rafael before helping design Apple's iPhone Portrait Mode.
- The $300 million acquisition merges Glass Imaging's optic software engineers directly into OpenAI's expanding hardware product group.
- Co-founder Ziv Attar brings extensive experience from Israeli defense R&D units and consumer optics engineering.
- The deal signals OpenAI's commitment to controlling custom camera capture capabilities for future physical AI hardware.
Startup
Enterprise urgency to block sensitive data leaks into AI workflows is driving explosive growth for specialized cybersecurity platforms — putting data loss prevention at the center of enterprise AI rollout.
BackgroundEnterprise adoption of generative AI tools increases data leakage risks when employees feed proprietary code or customer records into public language models. Data loss prevention systems monitor and govern data access pathways in real time.
- Existing investors Paladin Capital Group and YL Ventures participated in the round, bringing total capital raised to $112 million.
- MIND was founded by CEO Eran Barak, CTO Itai Schwarz, and VP R&D Hod Bin Noon to prevent sensitive corporate leaks.
- The company maintains headquarters in Seattle alongside its primary technical research center in Israel to support international scaling.
Startup
Buildots secured $130 million as tech hyperscalers rush to automate management of massive AI data center construction — establishing jobsite AI as core infra.
BackgroundBuildots uses wearable 360-degree helmet cameras paired with computer vision to compare physical jobsite progress against digital design blueprints. Rapid global AI expansion has created severe construction scheduling bottlenecks for data centers.
- The round brings total capital raised to $297 million, with participation from Lightspeed, Intel Capital, Qumra Capital, and Viola Growth.
- Enterprise clients including Intel, Digital Realty, Mortenson, and Hochtief use the platform as a digital construction control tower to prevent costly delays.
- Buildots plans to expand engineering operations to meet urgent project demands across data center and advanced manufacturing sites worldwide.
Startup
Keewano raised $12 million to replace relational tables with event-stream databases for AI agents — signaling how agent workloads are reshaping core database architecture.
BackgroundTraditional relational databases require complex transform pipelines and multi-table join operations to process user interaction histories. Event-driven architectures preserve raw activity streams, allowing machine learning models to analyze behavioral context directly.
- The funding round was led by Hetz Ventures, alongside Andreessen Horowitz (a16z Speedrun), Remagine Ventures, and DIG Ventures.
- KeewanoDB assigns isolated event tables per entity, executing behavioral queries across millions of records in under 0.5 seconds.
- Available as a managed service on Google Cloud, the database targets enterprise developers building autonomous AI agents.
Startup
Startup
Unlock the full brief
Sign in to read every signal, takeaway, and source. Free account — Apple, Google, or email.
