Singapore — In the global race to build the next generation of AI platforms, most startups rush to conquer Silicon Valley or target premium markets in North America and Europe. Bruce Yang, founder and CEO of Agnes AI, is taking a radically different approach: dominate Southeast Asia first, then expand outward.
It's a strategy that's already paying off. Since launching in July 2025, Agnes has amassed over 3 million registered users and 200,000 daily active users — with roughly half coming from Southeast Asia. The platform consistently ranks among the Top 10 productivity apps across Google Play stores in the Philippines, Vietnam, Indonesia, and other regional markets.
"We want to be the king of AI products in the region," Yang explained in a recent interview on the Convo AI World Podcast. "If we can achieve high penetration — maybe 50% or more — that gives us a baseline. Then we can copy the model to other regions like LATAM, the Middle East, parts of Europe, even Africa."
The growth metrics tell a compelling story. Agnes's customer acquisition cost is as low as 20 cents — a figure Yang says he hasn't seen matched by any comparable product. But what's more remarkable is that cheap acquisition hasn't come at the cost of retention.
"Usually, if acquisition is very cheap, retention is terrible," Yang noted. "People try it and disappear. But our retention is crazily good — day-30 retention is around 10%, on par with early Facebook. Week-8 retention is around 30%."
For context: Instagram took approximately 2.5 months to reach 1 million users. Snapchat took 10 months. Agnes reached 3 million in roughly two months — and users are staying.
Yang's regional-first strategy draws explicit inspiration from Southeast Asia's most successful tech companies. Just as Grab outcompeted Uber and Shopee and Lazada outmaneuvered Amazon in the region, Yang believes a locally-built AI platform can outperform global giants by understanding regional needs more deeply.
"In Southeast Asia, people feel pride when a strong product is built within the region," Yang said. "That positioning helps us grow. We support the culture, the language, the local context — things that global platforms often overlook."
Agnes supports minority languages commonly used across Southeast Asia that major global LLMs frequently ignore, delivering accurate understanding and generation tailored to local linguistic needs. The platform also offers significantly lower token costs than major LLM providers, making advanced AI accessible to users who cannot afford premium services.
Yang is candid about the lessons learned along the way. Early expansion into LATAM and the Middle East — while initially promising — revealed the limits of growing too fast without local presence.
"We made a mistake expanding too quickly," Yang admitted. "In Argentina and Mexico, our acquisition cost was even lower than Southeast Asia, and we quickly became #1 or #2 in some countries, even beating Gemini. But because we weren't locally present, we couldn't collect feedback effectively. Retention was much lower than Southeast Asia."
The experience crystallized the company's focus. "We stopped pushing aggressively outside the region. We're a small team — engineering and data science — and we have to grow cost-efficiently. Our strategy now: become the king of Southeast Asia first, then expand outward when we have a repeatable model."
Agnes positions itself as a "sovereign AI" company — one that builds and controls its own technology stack rather than depending entirely on overseas models and APIs. The company has developed a full family of proprietary models ranging from 7B to 30B parameters, including the state-of-the-art Agnes-R1.
This approach aligns with Singapore's broader NAIS 2.0 strategy, which aims to establish the city-state as a global hub for AI innovation and governance. Agnes exemplifies how locally-built, self-reliant AI systems can combine academic research, supportive policy, and commercial ambition.
"A lot of AI apps charge early because they don't run their own models — API costs are high," Yang explained. "We solve cost issues at the source by deploying our own models. That's how we can offer premium features for free while still scaling."
Agnes is currently in the process of closing a funding round worth tens of millions of U.S. dollars, at a valuation exceeding USD 100 million. Additional funding rounds, projected to value the company at USD 300–500 million, are expected to follow. The proceeds will fund the training of regional large language models and accelerate commercial expansion — but only when the Southeast Asian model is fully proven.
"If I'm able to grow like that, we have a very good chance to become an important global player in consumer AI in the next 2–5 years," Yang said. "But first, we need to win the region."
For a startup ecosystem often fixated on Silicon Valley validation, Agnes offers a different playbook: build where you're from, serve the users others overlook, and let regional dominance become the foundation for global expansion.
The AI-native Facebook may not be built in Menlo Park. It might be built in Singapore.