From approximately 80MW committed in 2026 to a projected 580MW by 2027, Aolani is making a major push to expand its AI infrastructure footprint in Malaysia! The addition of 500MW in AI data centre capacity is targeted to come online starting in the second half of 2027. These facilities will support its growing portfolio of AI factories and GPU deployments, serving global hyperscalers and AI-native companies. We believe the real opportunity extends beyond building more data centres. By making high-performance compute more accessible to local enterprises, researchers, and startups, Malaysia has an opportunity to become not just a hub for AI infrastructure, but also a place where more AI innovation is built. Full story here: malaymail.com/amp/news/money… @vishalharnal @khailee @AriefJohan @amirzakwan @atxfrench

🔥

AI keeps getting better at completing individual tasks. With Ampersand, @humansand is introducing a model designed to act as a collaborator, helping people work toward shared goals. Trained through simulations with Persimmon, humans&'s model of human behavior, Ampersand learns to consider the longer-term impact of its actions. In multi-month simulations, users working with Ampersand made greater progress toward their long-term goals than those using other frontier models evaluated. We're particularly interested in AI that understands not just what people ask for, but what they're ultimately trying to achieve. We look forward to seeing how developers bring this approach to new applications. Explore Ampersand here: ampersand.humansand.ai/blog @christine_tsai @TonyW @vishalharnal @khailee @dfect @atxfrench @leoclee

Think about how much faster robotics could advance if we didn't have to teach every new robot from scratch. That's the opportunity @dreamlabs_ai is pursuing. Founded by former NVIDIA robotics research leaders @jang_yoel, @GeYunhao, @scott_e_reed, and @youliangtan, the company has just emerged from stealth with US$90M in seed funding and the launch of Dream-1. Dream-1 learns from videos of humans performing tasks and applies that knowledge across robots from different manufacturers. What previously took months of programming could now take minutes. The company already works with 20+ hardware partners. And with Dream API now publicly available to developers worldwide, anyone can access this intelligence to turn camera feeds and plain-English instructions into robot actions, starting at US$3/hour. We look forward to seeing how Dream-1's capabilities expand further. Congratulations to the Dream Labs team! Start building at dreamlabs.ai. Full story here: businesswire.com/news/home/2… @christine_tsai @TonyW @vishalharnal @khailee @dfect @atxfrench @leoclee

Building more capable AI doesn't always have to mean throwing more compute at the problem. @Physion_Labs is demonstrating what's possible with NexTron, its AI training engine designed to make the process of training and fine-tuning large AI models faster and more memory-efficient, without requiring additional GPU hardware. In benchmarks, NexTron achieved up to 20.9× higher training throughput than Microsoft DeepSpeed ZeRO-3 on identical H100 hardware. The team also fine-tuned a 180B-parameter model using just eight H100 GPUs, with comparable model quality across ten benchmarks. We love seeing founders like Evelyn Qin Zhang and Bing Shuai challenge conventional thinking about how AI should advance. Physion Labs is a great example of why building smarter can be just as important as building bigger. Congratulations to the team on these incredible achievements! Read the full technical report: physionlabs.ai/blog/nextron @christine_tsai @TonyW @vishalharnal @khailee @dfect @atxfrench @leoclee

Elorian just launched Glimpse, a visual thinking model that reasons in images, not just about them. Frontier-level multimodal results at a fraction of the cost, built by a small team led by @AndrewDai & @yinfeiy 🚀 Live today via API + chat → elorian.ai/blog/introducing-… #500FounderStories

Apply before the deadline tonight → 500.co/founders/flagship

Congrats @ElorianAI on releasing Glimpse! ✨ Glimpse is a new category of visual thinking model: a system that reasons about the physical world from what it sees. On multimodal benchmarks, Glimpse attains frontier-level performance at the level of (+ sometimes exceeding) GPT-6 Astra and Claude Opus 5.5 while being substantially smaller & cheaper. The Elorian team built Glimpse w/ only a few hundred GPUs, post-trained on an open weights base model. @AndrewDai @yinfeiy @SethInternet @seojin_b Let's gooo!! 🚀🙌

Congratulations to Mark Lim, George-Cristian Potrivitu, and the Aliena team on raising US$5.5M in Series A1 funding! With sanctions disrupting Hall thruster supply chains and shortages delaying US defense satellite programs, reliable propulsion technology is becoming increasingly important. Aliena is positioning itself to meet this growing demand from satellite manufacturers and operators worldwide. With its own test and production facility in Singapore, the team can produce up to 100 propulsion systems annually. Aliena is targeting delivery times of 12 months, compared to the industry average of 12 to 18 months, with ambitions to reduce this to six months or less. We look forward to seeing more milestones from the team, including its expansion into Europe, where Aliena plans to establish an office by early 2027! Full story here: techinasia.com/sg-spacetech-… @vishalharnal @khailee @AriefJohan @atxfrench

Dapta has raised US$6M in seed funding to bring AI sales agents to more small businesses across the U.S. and Latin America, building on strong momentum with over 200K users, US$4M in annual revenue, and more than 4,000 new users joining weekly. Its AI sales agent, Dapti, instantly responds to inbound leads, qualifies prospects, and schedules meetings, helping SMBs convert more customers without adding administrative overhead. According to Dapta's internal data, businesses using Dapti are seeing their contact rates double. We're excited to continue backing Nicolas Rojas and the team as they make enterprise-level AI capabilities more accessible to small businesses. Full story here: globenewswire.com/news-relea… @christine_tsai @khailee @dfect @damarismendozal @rlomeli88 @DidierQuiroz @regina_zuru @porfin_entiendo @atxfrench

Congratulations StrongKeep on being named one of 18 finalists for the 2026 Emerging Enterprise Award! Recognizing young businesses that are finding innovative ways to address changing market needs, the award highlights exactly the kind of opportunity StrongKeep is pursuing: making cybersecurity accessible to smaller businesses that have traditionally been underserved. As cyber risks and regulatory requirements grow, StrongKeep is helping SMEs protect themselves without needing dedicated IT teams or technical expertise. Its AI-native platform bundles essential cybersecurity tools into one product, automating configuration and monitoring while helping businesses meet data protection and compliance requirements. Great to see Gaurav Keerthi and the team recognized for their work! Full story here: businesstimes.com.sg/singapo… @vishalharnal @khailee @AriefJohan @atxfrench @amirzakwan

Algolia has acquired Velou, a product intelligence company that uses multimodal AI to analyze product images and descriptions, automatically generating detailed attributes that help retailers match products to what shoppers actually search for. Better product data can translate directly into better search results and higher conversions. Velou brings a strong track record, including helping one retailer increase site search revenue by 60% and another boost search conversions by 33%. With the acquisition, @algolia can bring these capabilities to its 18,000+ customers, strengthening the product intelligence behind search, recommendations, personalization, and AI agents, without requiring retailers to rebuild existing integrations. Congratulations Stephen Lynch and the Algolia team! More on the acquisition: algolia.com/blog/algolia/alg… @christine_tsai @TonyW @vishalharnal @atxfrench @leoclee @dfect