Singapore’s deep tech startup SOMIN has earned international recognition after Google Deepmind cited its research in generative AI (GenAI) in a recent publication. The paper acknowledged SOMIN’s work as foundational to ACAI (AI Co-Creation for Advertising and Inspiration), a new system designed by DeepMind to automate ad creation and strategy development.

This milestone marks a significant moment for SOMIN, positioning the company as a key innovator in the evolving field of AI-powered marketing. While many generative AI tools focus on simplifying content production for users with limited creative experience, SOMIN has taken a more data-driven and strategic approach through its proprietary system, SOMONITOR.

At the heart of SOMIN’s contribution is SOMONITOR, an explainable AI framework designed for marketers to predict ad performance and refine strategies with clarity. Unlike many tools that act as black boxes, SOMONITOR provides transparent insights, allowing users to understand why particular decisions are made by the AI. This empowers marketing professionals to make more informed choices, tailoring content and delivery channels with precision.

One of SOMONITOR’s distinctive strengths lies in its predictive capabilities. It evaluates ads before they are launched, forecasting metrics such as click-through rates (CTR) using machine learning models trained on industry-specific data. This reduces the reliance on trial-and-error approaches, allowing teams to optimise campaigns with greater confidence.

In contrast to more template-driven AI tools such as ACAI or Mavic, which are typically aimed at non-specialist users, SOMONITOR serves as a robust platform for marketers seeking to enhance their strategic decision-making. By clustering content based on customer personas and aligning with real-time audience behaviours, the system facilitates more targeted and effective campaigns.

Also Read: Women in data: Busting myths, breaking barriers and building an inclusive future for tech

A growing concern in generative AI is the phenomenon of “hallucination”, where outputs appear convincing but are ultimately inaccurate. SOMIN addresses this challenge by anchoring its predictions in industry-verified data, ensuring recommendations remain reliable. Its integration with Global Web Index (GWI) also gives it access to rich consumer datasets, enabling deeper audience segmentation and more relevant content suggestions.

This data-backed approach is especially valuable as businesses navigate rapidly changing market conditions. Rather than relying solely on historical performance, SOMONITOR adapts to current consumer trends, an advantage that proves essential in dynamic sectors such as retail, automotive, and fast-moving consumer goods.

Collaborating with Google and the enterprise ecosystem

SOMIN’s recent recognition by Google DeepMind builds on its established relationship with Google. As an alumnus of the Google for Startups Accelerator and a certified Google Independent Software Vendor (ISV), SOMIN has benefitted from strategic partnerships, technical resources, and go-to-market support.

This includes receiving US$350,000 in Google Cloud credits and early access to the Gemini language model, bolstering SOMIN’s infrastructure and capabilities. Additionally, the company’s visibility within the Google ecosystem has led to industry connections with agencies such as Dentsu Asia Pacific, facilitating partnerships with brands including Toyota, Heineken, and Vienamil.

Through the Google Marketplace, enterprise customers can now directly access SOMIN’s solutions—an efficiency that streamlines procurement and expands reach.

Locally, SOMIN’s inclusion under Singapore’s Productivity Solutions Grant (PSG) and the IMDA GenAI Sandbox programme has lowered adoption barriers for small and medium-sized enterprises. By subsidising costs and promoting technological innovation, these initiatives have helped businesses implement AI without heavy upfront investments.

Several companies have reported substantial efficiency gains. Marketing agency NEO360, for example, reduced its proposal preparation time from three hours to 45 minutes, while BLAK Labs cut content planning hours by 70 per cent. Mothercare Singapore also improved its competitor research and customer segmentation workflows by as much as 80 per cent.

Also Read: Building an AI-ready Asia by bridging talent, technology, and cyber threats

These real-world outcomes demonstrate the utility of SOMIN’s solutions not just in theory but in daily operations across different industries.

Looking ahead: Human-AI collaboration in marketing

While large tech companies such as Meta and Google continue to dominate generative content production for SMEs, SOMIN is carving a niche in the enterprise segment where AI is increasingly seen as a collaborative tool rather than a replacement for human creativity.

By focusing on the four pillars of content strategy—social listening, competitor analysis, audience research, and business analytics—SOMIN is working to unify these domains under a single platform. Its long-term vision is to democratise strategic marketing insights, enabling businesses of all sizes to compete with data-backed decision-making.

—

Image Credit: SOMIN

The post Recognised by Google DeepMind, SOMIN aims to redefine AI-powered marketing appeared first on e27.