Vase AI

AI-powered consumer research platform giving companies access to 3.6M verified Southeast Asian consumers — insights in 24 hours

Most companies in Southeast Asia don't conduct market research — not because they don't want consumer insights, but because they believe research costs six figures and takes months to produce. By the time the results arrive, consumer behaviour has already shifted. Vase.ai was built specifically to dismantle that perception — and the reality behind it.

The real barrier isn't the price of research — it's the perception of it

Traditional market research agencies operate on a model designed for large companies with large budgets and long planning cycles. Commissioning a study means engaging a firm, scoping the methodology, fielding the survey, cleaning the data, and receiving a dense report weeks or months later. The price tag reflects this complexity — and for most SMEs and mid-market brands in Southeast Asia, that price tag places research in the "nice to have when we can afford it" category rather than the operating toolkit.

The irony is that the companies most in need of reliable consumer data are often the ones least able to access it. A local FMCG brand launching a new product, a D2C startup testing its positioning, a marketing team making a campaign allocation decision — these are exactly the contexts where fast, accurate consumer feedback would generate the highest return on investment. Instead, decisions get made on intuition, anecdote, and whatever data is available from internal analytics, which captures transactional behaviour but almost nothing about motivation, preference, or unmet need.

Vase.ai's founding insight, as articulated by CEO Julie Ng, was that this isn't a demand problem — it's a supply problem. Companies want to understand their consumers. What they don't want is a months-long research cycle that costs a fortune and delivers insights too stale to act on. Build something fast, affordable, and reliable, and the demand will follow.

The product: research-grade insights in 24 hours

Vase.ai's platform operates as a self-service research tool layered on top of a proprietary panel of 3.6 million identity-verified consumers across Southeast Asia. A brand manager or product team can log in, select their target audience from demographic, behavioural, and psychographic filters, use a pre-built expert template or write their own questions, and have responses streaming in within hours. The platform handles data cleaning, validation, visualisation, and AI-generated suggested insights — meaning the researcher receives a dashboard of actionable findings, not a raw data dump to interpret.

The 24-hour turnaround is not a marketing claim — it reflects the structural advantage of owning the panel rather than recruiting participants ad hoc for each study. When you have direct access to 3.6 million pre-verified consumers who are already part of the platform, fielding a survey to 400 nationally representative respondents takes hours, not weeks. That speed unlocks an entirely different use pattern: iterative research, where a product team runs a concept test, gets results the next morning, and uses those insights to inform the next sprint rather than the next quarter's planning cycle.

For organisations that want research expertise without building it internally, Vase.ai also offers managed service add-ons — survey scripting, full research reports — that maintain the speed advantage while adding the depth of a research agency. This hybrid model means Vase.ai serves both the self-sufficient research team and the brand marketer who knows what questions to ask but needs help designing the study.

The panel: the moat that most competitors can't replicate quickly

The Vase.ai Panel™ is the product's most defensible asset, and it's worth understanding specifically why. Consumer survey panels are only useful if they're trustworthy — and the market research industry has a well-documented problem with survey fraud: bots, duplicate responses, and paid survey mills that fill quotas with low-quality or fabricated answers. A fast survey with unreliable respondents produces confident but wrong conclusions, which is worse than no research at all.

Vase.ai's verification architecture addresses this with three independent layers: identity validation using government IDs verified via bank-level checks, device fingerprinting against cookies and IP metadata to detect duplicates, and mobile and email verification. Panel managers also conduct weekly manual reviews to remove suspicious respondents. The result is a panel where every response can be attributed to a verified individual, which is a materially higher standard than most self-service survey tools offer.

Building this panel took time, resources, and the kind of sustained community management that can't be shortcut. Vase.ai used referral mechanics as the primary acquisition channel — one of the most effective quality-preserving growth strategies for a panel, since referred respondents are more engaged and less likely to be fraudulent than those acquired through advertising. That panel is now the durable infrastructure that makes the 24-hour turnaround credible rather than merely aspirational.

The founder: a HKUST graduate who saw the research access gap directly

Julie Ng co-founded Vase.ai with CTO Asyrique Thevendran and COO Zhen Ng after identifying that the consumer research gap in Malaysia wasn't primarily a technology problem — it was a perception and access problem that technology could solve. Her background is in business and marketing, which gives her both the domain understanding of what research buyers need and the commercial instinct that has shaped the platform's go-to-market model.

The Digi-X corporate accelerator in 2019 provided the first external validation and RM335,000 in angel funding — a small amount by most standards, but sufficient for a team that was building a capital-efficient platform with a clear path to revenue. The subsequent funding round led by Indelible Ventures marked the transition from a locally-proven product to a regionally-scaled platform, with Indonesia expansion and deeper AI capabilities in the product roadmap. The clarity of the business model — a SaaS subscription layered on a proprietary data asset — means Vase.ai can scale revenue without proportionally scaling headcount.

What distinguishes the founding team is the combination of research domain depth and product-building discipline. The platform's AI insight engine was trained on years of expert researcher output — not simply a generic AI layer applied to survey data, but a model that reflects the judgment of experienced human researchers who know which correlations matter and which are artefacts. That training investment creates a compounding advantage: the platform gets smarter as more research is conducted through it.

What would have to be true for this not to work

The most direct competitive risk is platform commoditisation. Large global players — Qualtrics, SurveyMonkey, and increasingly Typeform — are all investing in AI-driven insights layers. If any of them builds or acquires a Southeast Asian panel of comparable quality, the differentiation shifts from the data asset to the platform experience and pricing. Vase.ai's defence is the depth of the SEA panel specifically — 3.6 million verified consumers with local identity validation is not something a global platform can replicate through a bolt-on acquisition.

The data quality challenge is ongoing. As the panel scales, maintaining the verification standards that make the data trustworthy requires sustained operational investment. A degradation in panel quality — even a perception of degradation — would undermine the core value proposition. The three-layer verification architecture and weekly manual review process are the right structural responses, but they must be maintained with discipline as volume grows.

The factors that drove conviction: a founding team that correctly identified the real barrier to market research adoption in SEA (perception and access, not demand), a proprietary data asset that took years to build and can't be replicated quickly, an AI insight layer trained on genuine research expertise rather than generic machine learning, and a pricing model that unlocks a market of companies that previously didn't do research at all. That combination — genuine data moat, relevant AI layer, and a radically expanded addressable market — is the kind of investment thesis that holds up at multiple scales.