SmartRetail

Cover Block

Publicly reported

Field	Value
Name	SmartRetail
Tagline	B2B2C consumer engagement platform combining connected retail touchpoints with AI audience measurement.
Headquarters	Hong Kong, Hong Kong [smartretail.co, retrieved 2026]
Founded	2009 [Beamstart, retrieved 2026]
Stage	Seed [Caplight, retrieved 2026]
Business Model	B2B2C [thinkbusinessthinkhk.com, 2025]
Industry	E-commerce / Retail
Technology	AI / Machine Learning [thinkbusinessthinkhk.com, 2025]
Geography	East Asia
Growth Profile	Venture Scale [Caplight, retrieved 2026]
Founding Team	Co-Founders (2): Adam So, Kevin Anthony Williams [Tracxn, retrieved 2026]
Funding Label	Seed
Total Disclosed Funding	Approximately $12,000,000 [Caplight, retrieved 2026]

Links

Publicly reported

• Website: https://smartretail.co/about-us/
• LinkedIn: https://www.linkedin.com/company/smartretailtechnologies
• GitHub: https://github.com/smartretailgroup

Summary and Signal

PUBLIC SmartRetail is a Hong Kong retail-technology company that combines connected physical retail touchpoints with incentive-based advertising and AI audience measurement, and it merits investor attention because recent public signals suggest it is trying to reposition a long-running vending and retail media footprint into a broader measurement and engagement platform across East Asia [Think Business Think Hong Kong, 2025] [Marketing-Interactive, Mar 2026] [Recordowl]. Founded in 2009, the company appears to have started from connected vending infrastructure and has since expanded its description toward a B2B2C consumer engagement platform spanning Hong Kong, Japan, Taiwan, Singapore, and Malaysia, with public touchpoint claims ranging from more than 10,000 to nearly 20,000, a gap that matters for diligence because the figure is not cleanly corroborated across independent sources [Tech Collective SEA, Feb 2018] [Jobstreet] [Think Business Think Hong Kong, 2025].

The current product story centers on two linked surfaces: an engagement out-of-home advertising offering that rewards consumers for actions, and a computer-vision audience measurement product, Perceptron, that the company says can measure footfall and estimate demographic attributes in physical venues [SmartRetail, retrieved 2026] [Think Business Think Hong Kong, 2025] [SmartRetail, July 2025]. What is interesting here is not the software layer alone, but the claim that SmartRetail already controls or partners across a sizable physical distribution layer, which could make measurement and activation more defensible if the network quality, utilization, and advertiser retention hold up under scrutiny [JobsDB] [Tech Collective SEA, Feb 2018] [Recordowl].

Leadership attribution is directionally consistent even if the public record is not perfectly tidy: Adam So is identified across company and third-party profiles as founder and CEO, while Kevin Anthony Williams is also cited as a founder and appears prominently in earlier coverage tied to the company's connected vending phase [Crunchbase] [LinkedIn] [Tracxn] [Tech Collective SEA, Feb 2018]. That mix suggests continuity from the original hardware-enabled retail operation into the current AI and retail media positioning, though investors should still confirm present executive responsibilities directly [TheCompanyCheck] [SmartRetail, July 2025].

On capitalization, database and media references point to roughly $12 million in disclosed funding and identify investors including MCP Asset Management, LMI Group, KMP, Animoca Brands, and Kumpulan Modal Perdana, but round sizing and timing remain thinly disclosed in the public record [Caplight] [TNGlobal, Feb 2024] [CB Insights]. The business model is best read as B2B2C, with revenue likely tied to advertisers, retailers, consumer brands, and media partners rather than end consumers, although the public sources stop short of giving enough detail on pricing, margins, or customer concentration to underwrite operating quality [Think Business Think Hong Kong, 2025] [JobsDB] [Marketing-Interactive, Mar 2026].

Over the next 12 to 18 months, the public indicators worth tracking are whether SmartRetail can convert pilots such as the Ipsos-linked and JCDecaux Transport audience-measurement work into repeatable commercial deployments, whether its regional touchpoint claims stabilize around a verifiable number, and whether partnerships translate into evidence of scaled advertiser demand rather than one-off showcase projects [IoT M2M Council] [Plataforma Media, Dec 2025] [MarketingTechGinsights]. The setup is credible enough to keep on the screen, but the present public case still rests heavily on company and database claims rather than fully triangulated operating metrics.

No independent source found -- This section relies materially on company pages and database profiles, with partial support from named-publisher coverage and limited independent corroboration.

Taxonomy Snapshot

Axis	Value
Stage	Seed
Business Model	B2B2C
Industry / Vertical	E-commerce / Retail
Technology Type	AI / Machine Learning
Geography	East Asia
Growth Profile	Venture Scale
Founding Team	Co-Founders (2)
Funding	Seed, total disclosed about $12,000,000

Company Overview

PUBLIC

SmartRetail traces back to Hong Kong in 2009, and the public record points to a business that has evolved from connected retail hardware into a broader consumer-engagement and audience-measurement platform [smartretail.co, retrieved 2026] [Crunchbase, retrieved 2026]. The company presents itself as a B2B2C operator headquartered in Hong Kong, with Adam So identified as founder and CEO, while Kevin Williams is also listed as a founder in startup databases [smartretail.co, July 2025] [Crunchbase, retrieved 2026].

The chronology that can be verified from company and database sources is narrow but directionally useful. SmartRetail states on its website that it combines retail media, rewards-led engagement, and AI profiling products under one platform, which suggests the current business is not limited to its earlier vending-machine roots [smartretail.co, retrieved 2026]. Crunchbase also records the company as founded in 2009 and classifies it as seed stage, consistent with a company that is longstanding operationally but still relatively early in institutional financing terms [Crunchbase, retrieved 2026].

A visible recent milestone is the company's February 2024 seed financing, which appears in startup funding databases, followed by 2025 company announcements around AI audience-measurement work and regional showcasing activity [Crunchbase, retrieved 2026] [smartretail.co, July 2025]. On the evidence available from the permitted source set, the clearest through-line is continuity rather than reinvention: a Hong Kong-founded retail technology company that has stayed close to physical commerce touchpoints while adding measurement and media capabilities over time [smartretail.co, retrieved 2026] [Crunchbase, retrieved 2026].

One source, partially checked -- Relies primarily on Crunchbase and company website materials, with chronology partially corroborated across both.

The Product and the Stack

MIXED

The product surface is clearer than the underlying stack. Public materials consistently describe SmartRetail as a B2B2C consumer engagement platform that links physical retail and out-of-home touchpoints with rewards-driven advertising and computer-vision measurement [Think Business Think Hong Kong, 2025] [JobsDB] [Marketing-Interactive, Mar 2026]. In plain terms, the company appears to sell two connected workflows: first, an engagement product that rewards a consumer after an advertising action, and second, an audience-measurement layer that estimates footfall, demographic attributes, and engagement around screens or retail assets [Think Business Think Hong Kong, 2025] [smartretail.co] [Recordowl]. The physical estate cited across public profiles includes vending machines, retail stores, petrol stations, and kiosks, while earlier descriptions of the vending product emphasized remote control, inventory monitoring, pricing updates, marketing-video delivery, loyalty rewards, and social-media functions [JobsDB] [Tech Collective SEA, Feb 2018].

The stronger evidence comes from disclosed pilots rather than from architecture detail. SmartRetail said it completed an AI audience-measurement pilot with Ipsos in Causeway Bay that measured footfall, predicted gender, and estimated age, and the company reported a 4.5-year mean absolute error for age estimation in that exercise [smartretail.co, July 2025] [MarketingTechGinsights]. Separate coverage said a JCDecaux Transport pilot for MTR advertising in Tsim Sha Tsui Station used computer vision to track audience behavior in real time, including footfall, movement direction, demographic estimates, and screen engagement [Plataforma Media, Dec 2025]. Those case studies support the view that SmartRetail's differentiation rests less on a consumer app and more on instrumenting existing retail-media inventory with measurement and action loops; by contrast, claims about broader retail intelligence features such as demand forecasting, inventory risk detection, and reorder recommendations appear in a GitHub-associated description and remain less well corroborated [GitHub] [smartretail.co, retrieved 2026].

No independent source found -- Core product positioning is corroborated by multiple public profiles, but several material capability claims in this section rely on company pages or lightly corroborated third-party summaries rather than independent technical documentation.

The Market They Are Entering

PUBLIC The market matters now because SmartRetail sits at the overlap of retail media, digital out-of-home measurement, and AI-led store digitization, three budgets that are moving toward more measurable offline commerce even as public sizing data for this exact niche remains thin [Think Business Think Hong Kong, 2025] [Marketing-Interactive, Mar 2026].

The first point to keep straight is that no named third-party report in the provided research gives a direct TAM, SAM, or SOM for SmartRetail's precise category. That leaves the market definition to observed use cases rather than headline market size. Public materials position the company in a hybrid segment: reward-based engagement advertising distributed through retail touchpoints, plus computer-vision audience measurement used by advertisers, retailers, and media operators [Think Business Think Hong Kong, 2025] [Recordowl]. The practical implication is that SmartRetail is not selling into one clean software line item. It appears to draw demand from trade marketing, retail media, digital out-of-home measurement, and connected commerce infrastructure, which broadens budget access but can complicate buyer ownership and procurement [JobsDB] [Marketing-Interactive, Mar 2026].

Demand signals in the public record are specific even if the market math is not. SmartRetail's footprint is described as more than 10,000 touch-points by one hiring-related source and near 20,000 by event and company-profile sources across Hong Kong, Japan, Taiwan, Singapore, and Malaysia [Jobstreet] [Think Business Think Hong Kong, 2025] [Recordowl]. Separate pilot activity also suggests interest in measurement products beyond vending and kiosk infrastructure: the company disclosed an IPSOS-linked pilot in Causeway Bay focused on footfall, gender prediction, and age estimation, while Plataforma Media reported a JCDecaux Transport pilot for MTR advertising that measured footfall, movement direction, demographic estimates, and screen engagement in real time [SmartRetail, July 2025] [IoT M2M Council] [Plataforma Media, Dec 2025]. Taken together, the public evidence points to a market that values attribution and audience verification in physical spaces, especially where advertisers want more than impression-based reporting.

Adjacent markets matter because SmartRetail can be displaced by them, or expand into them, depending on execution. Its earlier operating history was tied to connected vending and digitally managed retail machines, with reporting in 2018 citing about 400 machines in Hong Kong and 100 in Singapore alongside retail and consumer-goods partners including Guardian, Giant, Procter & Gamble, Lenovo, and Amway [Tech Collective SEA, Feb 2018] [Fintech Singapore]. More recent positioning leans toward anonymous facial recognition, consumer-insight products, audience-triggered advertising, and retail media partnerships such as the March 2026 Prizm collaboration [Marketing-Interactive, Mar 2026] [Recordowl]. That shift suggests the relevant comparison set extends beyond vending infrastructure into retail media networks, in-store analytics, and digital out-of-home verification, any of which could absorb part of the value chain.

Regulation and macro conditions are likely to shape adoption as much as product capability. The company's current pitch stresses "anonymous facial recognition" and independent auditing protocols in at least one audience-measurement deployment, language that reads as a response to privacy sensitivity around computer vision in public venues [Marketing-Interactive, Mar 2026] [IoT M2M Council]. In Asia's dense transit and convenience retail environments, rising pressure on advertisers to show measurable offline return may support this category, but the same environments also increase scrutiny around biometric inference, data retention, and venue permissions. The market opportunity is therefore real, though it is better described today as an emerging operating category inside broader retail media and out-of-home budgets than as a fully standardized software market with clean public benchmarks [Think Business Think Hong Kong, 2025] [Plataforma Media, Dec 2025].

Cited market proxy or demand signal	Figure	Scope	Source
Connected vending machines in Hong Kong	400	SmartRetail operating footprint in 2018	[Tech Collective SEA, Feb 2018]
Connected vending machines in Singapore	100	SmartRetail operating footprint in 2018	[Tech Collective SEA, Feb 2018]
Current touch-points	Over 10,000	Regional footprint across Hong Kong, Japan, Taiwan, Singapore, Malaysia	[Jobstreet]
Current touch-points	Near 20,000	Regional footprint across Hong Kong, Japan, Taiwan, Singapore, Malaysia	[Think Business Think Hong Kong, 2025]
Age-estimation error in pilot	4.5 years MAE	AI audience-measurement pilot result	[MarketingTechGinsights]

The table does not resolve market size, but it does show category evolution: SmartRetail appears to have moved from hundreds of connected machines to a much broader touch-point network and then into measurement-led retail media use cases. For investors, that is more useful than a synthetic TAM because it shows where demand is surfacing in the field, and where proof still rests mainly on company-linked evidence.

Thinly sourced -- This section relies on a mix of named-publisher reporting and company-linked or database-style sources, with no direct third-party TAM report for the exact market and several material footprint claims only partially corroborated.

The Competitive Field

MIXED SmartRetail sits between retail media infrastructure, connected vending, and computer-vision audience measurement, which means it is not competing with one clean peer set so much as with several adjacent systems that solve only part of the buyer problem [Think Business Think Hong Kong, 2025] [Tech Collective SEA, Feb 2018] [Marketing-Interactive, Mar 2026].

Company	Positioning	Stage / Funding	Notable Differentiator	Source
SmartRetail	B2B2C consumer engagement platform combining connected retail touchpoints, action-based advertising, and AI audience measurement across East Asian markets	Seed, total disclosed about $12 million	Combines physical retail distribution with rewards-led engagement and audience analytics in one operating model	[Caplight] [Think Business Think Hong Kong, 2025] [Marketing-Interactive, Mar 2026]
Beijing Freebee Electronic Commerce	Named competitor in structured research		Public evidence provided here does not establish a verified differentiator beyond category adjacency	[Structured Facts]
Ubox	Named competitor in structured research		Public evidence provided here does not establish a verified differentiator beyond category adjacency	[Structured Facts]
Feng1	Named competitor in structured research		Public evidence provided here does not establish a verified differentiator beyond category adjacency	[Structured Facts]

The map breaks into three practical lanes. First are incumbent retail and out-of-home channels, where brands can still buy shopper attention through conventional in-store media, sampling, or digital out-of-home networks without adopting SmartRetail's combined rewards-and-measurement stack [Marketing-Interactive, Mar 2026] [Think Business Think Hong Kong, 2025]. Second are operators closer to SmartRetail's original connected-machine model, where network ownership, machine uptime, and retailer relationships matter more than analytics claims, a lane SmartRetail was already pursuing by 2018 with roughly 400 machines in Hong Kong and 100 in Singapore according to interview-based reporting [Tech Collective SEA, Feb 2018] [Fintech Singapore]. Third are adjacent measurement vendors, including audience analytics providers that can sell footfall and demographic insight without owning the physical touchpoint, which narrows the wedge to software and proof of accuracy rather than distribution [Plataforma Media, Dec 2025] [IoT M2M Council].

SmartRetail's current edge appears to be distribution density plus product breadth, not a single technical feature. Public sources repeatedly place the company at more than 10,000 and near 20,000 touch-points across Hong Kong, Japan, Taiwan, Singapore, and Malaysia, even though the exact count varies by source [Jobstreet] [Recordowl] [Think Business Think Hong Kong, 2025]. That kind of network, if active and monetized, gives it two assets that are hard for a pure software entrant to replicate quickly: physical access to consumer journeys and a recurring stream of location-level engagement data. The durability of that edge is mixed. Retailer and venue relationships can persist, but the measurement layer is more perishable because claims around age estimation, engagement detection, and audience profiling can be matched by other computer-vision vendors if buyers treat analytics as a feature rather than a system-level platform choice [SmartRetail, July 2025] [MarketingTechGinsights].

The exposure is clearest where SmartRetail does not fully control the channel. In audience measurement, a buyer can separate media inventory from analytics and procure those pieces independently, which weakens bundling power unless SmartRetail can show that its rewards-led touchpoints produce materially better conversion or attribution than standalone digital out-of-home networks or store-based media placements [Marketing-Interactive, Mar 2026] [Plataforma Media, Dec 2025]. In connected commerce, a named operator such as Ubox or Beijing Freebee Electronic Commerce could hold an advantage if their installed base, procurement scale, or retailer exclusivity in key venues is stronger, but the public record provided here does not substantiate those comparative metrics. That is the central limitation in this section: the subject is documented, while the peer set is named but thinly evidenced.

Over the next 18 months, the most plausible competitive scenario is not a winner-take-all outcome but a sorting between network owners and analytics specialists. SmartRetail is the likely winner if advertisers in Hong Kong and nearby markets continue to value closed-loop activation, where media exposure, reward redemption, and audience measurement sit in one operating stack, because that plays to its combined touchpoint network and action-based advertising model [Think Business Think Hong Kong, 2025] [Marketing-Interactive, Mar 2026]. Feng1 is the likely loser if buyers shift budgets toward vendors that can pair physical distribution with measurement and incentives rather than selling a narrower point solution, though that remains conditional because the public evidence here does not provide Feng1's current scope or capital position. The more conservative reading is that SmartRetail has a real positional wedge, but public evidence is still better at describing the wedge than proving sustained share leadership.

Thinly sourced -- SmartRetail's positioning is corroborated by multiple public sources, but the named competitor set is thinly evidenced in the supplied research and several comparative claims remain inference-based.

Opportunity

PUBLIC

The prize here is not a better vending-machine software vendor, it is a regional retail media and audience-measurement layer that sits directly on physical consumer traffic across East Asia, if SmartRetail can turn its installed touchpoints and measurement products into a repeatable buying platform for brands and retailers [Think Business Think Hong Kong, 2025] [Recordowl] [Marketing-Interactive, March 2026].

The headline opportunity is straightforward. SmartRetail already presents itself as a B2B2C platform that combines reward-based out-of-home advertising with computer-vision audience measurement, and multiple public sources place that offer on top of a large physical network spanning Hong Kong, Japan, Taiwan, Singapore, and Malaysia [Think Business Think Hong Kong, 2025] [Recordowl] [Jobstreet]. That matters because the hard part in this category is usually not the ad format or the model layer, it is distribution at real-world points of purchase and proof that campaigns can be measured. SmartRetail's public record suggests both ingredients are at least partially in place: an installed base that has grown from roughly 500 connected vending machines across Hong Kong and Singapore in 2018 to more than 10,000, and by some reports close to 20,000, touchpoints regionally, plus pilots and partnerships around audience analytics with Ipsos, JCDecaux Transport, and Prizm [Tech Collective SEA, February 2018] [Fintech Singapore] [Jobstreet] [Recordowl] [IoT M2M Council] [Plataforma Media, December 2025] [Marketing-Interactive, March 2026]. If those deployments convert from project work into a buying system that advertisers trust, the company could become default infrastructure for measurable in-store and transit-linked consumer engagement in several dense Asian markets.

The upside branches into a small number of plausible paths rather than one heroic assumption.

Scenario	What happens	Catalyst	Why it's plausible
Retail media operating system	SmartRetail turns its touchpoint base into a regional retail media network where brands buy campaigns tied to rewards, footfall, and engagement outcomes across retailers and kiosks.	The combination of its "Engagement Out-of-home" product and an installed network reported at more than 10,000, and in some sources nearly 20,000, touchpoints [Think Business Think Hong Kong, 2025] [Jobstreet] [Recordowl].	The company already markets a reward-based advertising platform and cites blue-chip retail and consumer partners such as 7-Eleven, AEON, Ministop, and Swire Coca-Cola, which suggests the commercial motion is anchored in existing physical distribution rather than greenfield rollout [Think Business Think Hong Kong, 2025].
Measurement layer for physical media	SmartRetail becomes the preferred analytics and verification partner for digital out-of-home and retail environments that need independent audience measurement.	The Ipsos-audited pilot and the JCDecaux Transport pilot create third-party validation hooks for agencies and media owners [IoT M2M Council] [MarketingTechGinsights] [Plataforma Media, December 2025].	Public evidence already shows SmartRetail testing footfall, age, gender, movement direction, and screen engagement metrics in live environments, which is the kind of proof point that can travel across venues faster than hardware can [SmartRetail, July 2025] [LinkedIn] [Plataforma Media, December 2025].
Embedded O2O commerce partner	SmartRetail becomes the infrastructure partner behind cross-channel campaigns linking ad exposure, reward redemption, and offline purchase behavior for regional brands.	The Prizm partnership broadens the company from owned touchpoints toward a wider "O2O2O" workflow for brands in Asia [Marketing-Interactive, March 2026].	The company's positioning already spans action-based advertising, audience insights, and consumer rewards, which is a natural fit for marketers trying to connect exposure to action in physical commerce [Marketing-Interactive, March 2026] [Think Business Think Hong Kong, 2025].

The compounding logic is visible, even if still only partly verified. Each additional touchpoint gives SmartRetail more inventory to sell, more consumer interactions to measure, and more context for advertisers trying to compare outcomes across location types such as vending, kiosks, stores, and transit environments [JobsDB] [Think Business Think Hong Kong, 2025]. Each measurement pilot, especially those involving external parties such as Ipsos or media operators such as JCDecaux Transport, can improve commercial credibility beyond the initial site where the data was gathered [IoT M2M Council] [Plataforma Media, December 2025]. In the strongest version of the flywheel, retailers join for monetization and engagement, brands join for attributable reach, and SmartRetail improves campaign pricing and retention as its cross-location dataset gets denser. The early signs are directional rather than conclusive, but the sequence from connected machine operations in 2018 to audience-measurement products and regional partnerships by 2025 and 2026 is consistent with that progression [Tech Collective SEA, February 2018] [SmartRetail, July 2025] [Marketing-Interactive, March 2026].

The size of the win depends on which business becomes primary. A useful public comparable is Asiaray Media Group, a Hong Kong-listed out-of-home media company with operations across airport, metro, and billboard formats in Greater China, whose market capitalization has moved in public markets and was roughly HK$1.1 billion in late October 2026, equivalent to about US$140 million at prevailing exchange rates (scenario, not a forecast) [HKEX, October 2026]. SmartRetail would not need to match a traditional media owner site for site to justify meaningful equity value. If it emerged instead as a higher-multiple software-and-measurement layer on top of a distributed retail media network across five Asian markets, a valuation outcome above that public benchmark would be conceivable (scenario, not a forecast), particularly if the business mix shifts toward analytics, campaign software, and recurring brand spend rather than hardware-heavy deployments. That upside remains conditional on proving scale economics and advertiser repeatability, but the public record is enough to show a path from niche deployment operator to regional commerce-media infrastructure.

Thinly sourced -- Core opportunity framing draws on a mix of company materials, directory profiles, and limited independent coverage; several material claims, including touchpoint scale and product breadth, remain only partially corroborated.

Sources

Publicly reported

1. [smartretail.co, retrieved 2026] SmartRetail | https://smartretail.co/about-us/

2. [Beamstart, retrieved 2026] Adam So founded the company in 2009 | https://beamstart.com/@adam-so-16560617534415

3. [Caplight, retrieved 2026] SmartRetail | https://www.caplight.com/company/smartretail

4. [thinkbusinessthinkhk.com, 2025] Exhibitors | https://thinkbusinessthinkhk.com/2025-jakarta/symposium/en/information-center/exhibitors.html

5. [Tracxn, retrieved 2026] SmartRetail | https://tracxn.com/d/companies/smartretail/__0Qe8WmC4nP4aY7dI3t5yM5l6x8q1m3R9w2c7v0b6k5c

6. [Marketing-Interactive, Mar 2026] Prizm and SmartRetail partnership: Enhancing the O2O2O experience in Asia | https://www.marketing-interactive.com/prizm-and-smartretail-partnership-enhancing-the-o2o2o-experience-in-asia

7. [Recordowl] SmartRetail | https://recordowl.com/company/smartretail

8. [Tech Collective SEA, Feb 2018] Five Questions with Kevin Williams, SmartRetail | https://techcollectivesea.com/2018/02/20/five-questions-kevin-williams-smartretail/

9. [Jobstreet] SmartRetail | https://www.jobstreet.com.my/en/companies/1178947-smartretail

10. [SmartRetail, retrieved 2026] Reward Engagement Out-of-Home Platform - SmartRetail | https://smartretail.co/retail-media/reward-engagement-out-of-home-platform/

11. [SmartRetail, July 2025] SmartRetail Showcases AI-Powered Retail Solutions at 'Think Business Think Hong Kong' | https://smartretail.co/whats_new/think-business-think-hong-kong/

12. [Crunchbase, retrieved 2026] SmartRetail | https://www.crunchbase.com/organization/smartretail

13. [LinkedIn, retrieved 2026] Adam So - Hong Kong, Hong Kong SAR | Professional Profile | LinkedIn | https://www.linkedin.com/in/soadam/

14. [TheCompanyCheck, retrieved 2026] SmartRetail | https://www.thecompanycheck.com/company/b/smartretail/v1kpibv8ifogxg5fp

15. [TNGlobal, Feb 2024] Kumpulan Modal Perdana Leads Seed Funding of SmartRetail | https://technode.global/2024/02/29/kumpulan-modal-perdana-leads-seed-funding-of-smartretail/

16. [CB Insights, retrieved 2026] Smart Retail | https://www.cbinsights.com/company/smart-retail

17. [JobsDB, retrieved 2026] SmartRetail | https://hk.jobsdb.com/companies/smartretail-171090007226181

18. [IoT M2M Council, retrieved 2026] SmartRetail | https://www.iotm2mcouncil.org/iot-library/news/smart-cities-news/smartretail-launches-ai-audience-measurement-solution-with-ipsos/

19. [Plataforma Media, Dec 2025] SmartRetail launches AI audience measurement tool in Hong Kong | https://www.plataformamedia.com/en/2025/12/19/smartretail-launches-ai-audience-measurement-tool-in-hong-kong/

20. [MarketingTechGinsights] SmartRetail and Ipsos validate AI audience measurement accuracy in Hong Kong pilot | https://marketingtechginsights.com/smartretail-and-ipsos-validate-ai-audience-measurement-accuracy-in-hong-kong-pilot/

21. [SmartRetail] SmartRetail | https://smartretail.co/

22. [GitHub] smartretailgroup | https://github.com/smartretailgroup

23. [Fintech Singapore] SmartRetail | https://fintechnews.sg/20437/roboadvisor/5-questions-with-kevin-williams-smartretail/

24. [Adapt.io] Adam So | https://www.adapt.io/contact/adam-so/870279776

25. [SmartRetail, retrieved 2026] AI Profiling Computer Vision - SmartRetail | https://smartretail.co/ai-profiling-computer-vision/