AI Shopping: what are the top startups now?d
In our AI shopping market deck, you will find everything you need to understand the market
SUMMARY
AI Shopping: what are the top startups now? The strongest names right now are Phia, ShopMy, Constructor, Gensmo, Daydream, Onton, Lily AI, Cernel, ShopOS, Alta, and Tilt, but they are winning in different parts of the market.
The most important pattern is that AI shopping is not one category anymore. It is splitting into consumer apps, creator commerce, retailer infrastructure, product data, visual search, live selling, and agentic checkout.
Phia looks like the clearest consumer breakout because it combines three signals that rarely appear together in early consumer AI: 1.5 million users, 9,600 brand partners, and a simple price-confidence use case.
ShopMy may be the stronger company if we define AI shopping more broadly. Its creator network, transaction flow, profitability, and $1.5 billion valuation show commercial proof most AI shopping assistants still do not have.
The best-funded startup is not automatically the best-proven one. Gensmo’s $60 million-plus seed round is a serious signal, but Phia’s funding came after visible consumer adoption, which makes the evidence cleaner.
Fashion AI still has a habit problem. Daydream has catalog breadth, Gensmo has ambition, and Alta has a smarter personal-data loop, but none has the same public usage proof as Phia.
The stronger wedges are narrow and practical. Price confidence, resale comparison, visual decision-making, product attributes, seller listing automation, and agent-ready data all feel more defensible than a generic “ask our AI what to buy” chatbot.
Merchant infrastructure may become more valuable than consumer-facing agents. Constructor, Lily AI, Cernel, and ShopOS sit closer to the data, ranking, and operations layers retailers need before AI shopping can work reliably.
Agentic checkout changes the competitive map. If users buy inside Perplexity, ChatGPT, Google, Amazon, or retailer-owned assistants, standalone startups need a sharper role before or around checkout.
The market already shows a useful divide between trust and automation. ShopMy owns human taste at scale, while Phia, Constructor, Lily AI, Cernel, and ShopOS help automate confidence, discovery, data, and operations.
The weakest zone is broad AI shopping assistants with no disclosed user base, merchant network, conversion lift, transaction flow, or proprietary data loop. Those companies may still become interesting, but today they do not deserve top ranking.
So the clean conclusion is that AI shopping has no single winner yet. Phia leads consumer shopping confidence, ShopMy leads creator-led commerce, and Constructor, Lily AI, Cernel, and ShopOS are building the infrastructure that retailers and agents will need.
This market map, featured in our AI shopping market deck, highlights top companies and startups in the AI shopping market
Which AI shopping apps are people actually using now?
Phia, Onton, Daydream, Gensmo, and Alta are the consumer AI shopping apps worth watching now, but Phia is clearly the strongest current breakout.
Phia stands above the pack because it has the rarest combination in consumer AI commerce: fresh user growth, merchant coverage, and a simple reason to come back. By late May 2026, Phia reported 1.5 million users and 9,600 brand partners, after launching in April 2025. That is stronger evidence than a launch waitlist or a polished demo because it shows real consumer pull across a large retail graph.
The reason Phia is ahead is also product-specific. It starts with a very concrete shopping question: “Am I buying this at the right price?” That is easier to adopt than a full AI fashion agent because price comparison, resale value, price tracking, and alternatives are immediate utilities.
Onton is the second-best usage story. It reported growth from around 50,000 to more than 2 million monthly active shoppers by late 2025. That is a bigger absolute active-user claim than most startups in the category disclose, but the context matters: Onton started in furniture and home, where visual shopping pain is very real. The open question is whether it can carry that behavior into apparel and consumer electronics with the same intensity.
Daydream is strong on ambition and catalog depth, but weaker on visible habit. Its June 2025 launch reportedly covered more than 2 million products from around 8,500 brands. That gives shoppers a lot to search through, but today we still do not have the kind of public usage metric that lets us place it above Phia or Onton. So Daydream is a credible platform candidate, while Phia is the clearer consumer breakout.
Gensmo sits in a different bucket. Its $60 million-plus seed funding in 2025 makes it one of the most heavily backed AI shopping startups, but funding is still a proxy. Until we see public user numbers, Gensmo is best read as a high-conviction investor bet on AI fashion discovery, not yet as a proven consumer winner.
Alta is smaller but sharper. Its June 2025 $11 million seed round backed a product built around wardrobe memory, receipts, styling, and avatar-based outfit planning. That is less broad than Daydream and less proven than Phia, but the thesis is smarter than generic AI search: shopping gets better when the assistant knows what you already own.
If you want more recent data on this point, please see our latest AI shopping market report.
Which AI shopping startups look strongest when we compare funding to proof?
Phia, ShopMy, Gensmo, Constructor, and ShopOS stand out, but not for the same reason.
Phia has the best funding-to-usage balance among pure AI shopping apps. Its $35.5 million round in May 2026 came after visible user growth, and not before. That sequencing matters. In this market, a company that raises after proving consumer adoption deserves more weight than a company that raises mainly on the promise of what AI shopping could become.
Gensmo has the biggest early-stage funding signal, with more than $60 million disclosed in 2025. That is impressive, but compared with Phia, the proof is more investor-side than user-side. We should treat it as a serious contender because elite investors are underwriting a large AI fashion-agent vision, while also being clear that the public traction bar is not yet the same.
ShopMy is the strongest company if we widen the definition from “AI shopping assistant” to “AI-era shopping infrastructure.” In October 2025, it raised $70 million at a $1.5 billion valuation. More importantly, it had already shown operating strength: 200% revenue growth over the prior year, profitability since 2024, 243,000 creators, and more than $200 million in monthly fashion, jewelry, and homeware sales reportedly flowing through the platform. That is much stronger commercial proof than most AI shopping startups can show today.
Constructor is the mature B2B benchmark. Its June 2024 round valued it at $550 million, and PitchBook lists total funding around $131 million. Compared with consumer apps, Constructor is less exciting culturally, but its business is closer to the money: retailers already pay for search, ranking, recommendations, and product discovery.
ShopOS is the emerging infrastructure name. Its $20 million June 2025 round is meaningful because it targets the operational layer for ecommerce brands: listings, campaigns, and store adaptation. The reason we rank it below Constructor and ShopMy is simple: it has strong founding and funding signals, but not yet the same public customer or revenue proof.
As this chart shows, and as featured in our AI shopping market deck, search interest in AI shopping has grown significantly
Which fashion AI startups look real, not just well-funded?
Phia, Daydream, Gensmo, Alta, shoppin’, and BNTO are the strongest fashion AI shopping group, with Phia and ShopMy-like trust models setting the bar higher than pure discovery apps.
Phia leads because it solves a narrower fashion problem better than most broad assistants. Fashion discovery is subjective, but price confidence is universal. If a shopper sees a jacket, Phia can help compare retail, resale, and alternatives. That makes it useful even when the AI does not fully understand taste.
Daydream is the most credible attempt to build a large fashion-search destination. Its launch catalog of roughly 2 million products and 8,500 brands gives it more breadth than most early fashion assistants. The weakness is habit. Today, a new shopping destination has to fight TikTok, Google, Amazon, retailer apps, ChatGPT, and Perplexity at once. Catalog size helps, but it does not automatically create a daily behavior.
Gensmo is the biggest “AI stylist as platform” bet. The $60 million-plus seed round gives it more room to build heavy product, visual, and personalization infrastructure than smaller rivals. But compared with Phia, it has less public evidence that consumers are already using it at scale. That is why we rank Gensmo as a high-upside contender rather than the current category leader.
Alta is probably the most interesting personal-fashion product. The wardrobe-memory angle gives it a better data loop than generic product search. If users upload their closet, receipts, and preferences, the assistant can recommend outfits with more context. That does not prove mass adoption, but it gives Alta a clearer reason to exist than another “find me a dress” chatbot.
shoppin’ is early, but the direction feels very current. Its $3 million seed round is small next to Gensmo, yet the product is aimed at prompt, image, mood, and social-inspiration discovery. For Gen Z fashion, that may be closer to how shopping actually starts these days: with a vibe, a screenshot, a creator, or a moodboard.
BNTO is also worth watching because it connects AI styling to rental, resale, and new purchase options. That is a stronger commerce model than styling alone. The question is whether it can build enough supply and user frequency to matter beyond a niche audience.
If you want more recent data on this point, please see our latest AI shopping market report.
Which startups are best at solving shopping confidence?
Phia is the clear leader here, while Onton and Zellor are the two smaller names that make the most sense.
This is where Phia looks most differentiated. Many AI shopping apps try to help users discover more products. Phia helps users decide whether a product is worth buying. That is a better wedge because shopping confidence is tied to money, regret, resale value, and timing. As seen above, its reported user and brand-partner scale makes that wedge feel more real than just a feature idea.
Onton solves confidence through visual context. Furniture and home shopping are full of uncertainty: size, style, fit, room compatibility, and alternatives. Onton’s infinite-canvas and generated-visual approach attacks that problem directly. Compared with Phia, it is less about price and more about imagination. That gives it a strong home-shopping angle, though the proof outside home is still thinner.
Zellor is much earlier, but it has one concrete merchant-side proof point: in April 2026, it raised €850,000 and reported client conversion lifts of up to 25% from its AI video shopping assistant. That is not enough to call it a market leader, but it is exactly the kind of metric we want from a young startup. A conversion lift, even if early and client-specific, is more useful than another vague claim about “personalized shopping.”
So in this category, Phia is the consumer-confidence leader, Onton is the visual-confidence challenger, and Zellor is the small but practical conversion tool to track.
This chart, featured in our AI shopping market deck, illustrates yearly VC funding for AI shopping startups
Which startups are building the pipes behind AI shopping?
Constructor, Lily AI, Cernel, ShopOS, and Athos Commerce are the key infrastructure names, with Cernel the freshest one to watch.
Constructor is the safest infrastructure pick because retailers already understand the value of search and product discovery. AI shopping agents can only recommend what they can interpret, rank, and retrieve. Constructor already sits in that workflow. Compared with new agentic-commerce startups, it has less narrative freshness, but much stronger enterprise maturity.
Lily AI is important because product data is becoming the hidden battlefield. Its public materials mention a 25,000-plus customer-oriented attribute taxonomy, more than 30 million items tagged per year, and over 3 billion training data points. That matters because shopping agents need attributes like fit, occasion, material, style, and intent, not just SKU names. Compared with Constructor, Lily is less about ranking results and more about making the product understandable in the first place.
Cernel is the emerging name that feels most “right now.” In February 2026, it raised €4 million in seed funding to build data infrastructure for agentic commerce. That is small compared with Constructor or ShopOS, but the timing is excellent. If AI agents become shopping intermediaries, merchants will need to structure product data for machines, not only for human browsing pages.
ShopOS is broader and more operational. Its June 2025 $20 million round positions it around AI agents for ecommerce operations: product listings, marketing campaigns, and store adaptation. Compared with Cernel, it is less narrowly focused on agentic product data, but it could matter more to brands if it becomes their everyday AI operations layer.
Athos Commerce is more of a consolidation signal than a fresh startup story. The January 2025 combination of Klevu and Searchspring brought together AI search and discovery tools used by thousands of brands. The takeaway is that older ecommerce search companies are moving fast to reframe themselves for the AI shopping era. That raises the bar for younger infrastructure startups: they need a real wedge, not just “AI search” language.
If you want more recent data on this point, please see our latest AI shopping market report.
Which startups are strongest in visual search and try-on?
Gensmo, Alta, Onton, shoppin’, and BNTO are the strongest visual-shopping group, but they attack different moments in the journey.
Gensmo is the most ambitious technical bet. Its work around AI fashion discovery, personalization, and virtual try-on puts it closest to the “AI stylist” vision investors love. The funding gives it room to build complex visual features that smaller startups may struggle to finance. But compared with Onton or Phia, we have less public evidence that shoppers already use it repeatedly.
Alta has the better personal-data loop. Virtual try-on can be gimmicky when it starts from a random product catalog. Alta’s stronger angle is that it starts from the user’s own wardrobe and avatar. That makes the styling problem more grounded. It does not need to show every product in the world. It’s simpler than that: it needs to make better use of what the user owns and may buy next.
Onton is the strongest visual-decision startup, especially for home. Its product actually helps shoppers imagine combinations and contexts. That matters because visual shopping is often about reducing uncertainty before purchase. Compared with Gensmo, Onton has stronger disclosed usage momentum; compared with Alta, it is less personal but more discovery-oriented.
shoppin’ is the cultural search bet. Prompt, screenshot, mood, and social-inspired search feels more natural for fashion than a classic keyword box. The company is earlier than the others, but if visual shopping keeps moving toward TikTok-like discovery, this kind of interface becomes more important.
BNTO rounds out the group because its AI styling connects to multiple transaction paths: rent, buy new, or buy secondhand. That is a useful distinction. A good styling engine should not only say “this looks good” but actually route the user toward the most sensible way to access the item.
This chart, featured in our AI shopping market deck, shows why Constructor is winning in AI shopping
Which startups are turning AI shopping into social commerce?
ShopMy, Tilt, Phia, and BNTO are the names that matter most here, with ShopMy far ahead on commercial proof.
ShopMy is the obvious leader. It is not a pure AI assistant, but it is one of the strongest shopping companies in the AI era because it owns trust at scale. In 2025, it reported or was reported to have 243,000 creators, more than $200 million in monthly sales flowing through the platform, 200% revenue growth, and profitability since 2024. Compared with most AI shopping apps, that is a different league of monetization proof.
The deeper point is that AI may make product discovery cheaper, but it does not automatically make taste trustworthy. ShopMy sits exactly in that gap. Creators tell people what to buy, and AI can later help organize, attribute, optimize, and personalize that flow.
Tilt is the live-shopping startup to watch now. In June 2026, it raised $26 million, bringing total funding to $50 million. Its AI features are practical: sellers can use Snap to create auction listings from camera input, generate descriptions, and get suggested pricing. Compared with consumer shopping agents, Tilt applies AI where friction is obvious: helping sellers list faster and sell live.
Phia has a lighter social angle, mostly through cultural distribution. Its investor base includes celebrities and fashion-adjacent names, which is not proof by itself. But in fashion shopping, cultural trust can lower acquisition costs. That makes it more relevant than it would be in a purely technical B2B market.
BNTO is the smaller social-commerce-adjacent bet. Its mix of styling, rental, resale, and purchase options could work well in taste-based communities, but it still needs clearer scale signals before we can rank it near ShopMy or Tilt.
Which startups could win if AI agents start checking out for us?
Perplexity, Cernel, Constructor, ShopOS, and Phia are the best-positioned names, though only some are pure startups in AI shopping.
Perplexity matters because it moved shopping inside an answer engine. Its November 2024 shopping launch and “Buy with Pro” feature showed what agentic commerce could look like: search, answer, recommendation, and checkout collapsing into one interface. That makes Perplexity a dangerous reference point for every AI shopping startup. If users can buy inside a general AI app, standalone shopping agents need a sharper reason to exist.
Cernel is the young startup most directly built for that shift. If agents are going to compare products on behalf of users, merchants need data that agents can read properly. That includes attributes, availability, compatibility, prices, and context. Cernel’s February 2026 seed round is small, but the company is aiming at a problem that gets bigger if agentic checkout becomes normal.
Constructor benefits from the retailer side of the same trend. Retailers will not want all AI shopping decisions to happen inside OpenAI, Google, Perplexity, or Amazon. They need their own discovery layer to stay competitive. Constructor already sells into that need.
ShopOS could matter if brands need to adapt their operations for agentic surfaces. If product listings, campaigns, and storefronts have to be reworked for AI intermediaries, automation becomes valuable. The company is early, but the workflow pain is plausible and current.
Phia is the consumer-side bridge. Even if checkout moves into AI agents, users may still want a neutral layer that checks price, resale value, and alternatives before purchase. That gives Phia a possible role upstream of checkout rather than inside checkout itself.
If you want more recent data on this point, please see our latest AI shopping market report.
This chart, featured in our AI shopping market deck, illustrates yearly funding for AI shopping startups
Which AI shopping categories still look too vague to crown winners?
Generic AI shopping chatbots still look weak unless they show usage, merchant coverage, conversion lift, or repeat behavior.
This is where we need to be strict. A startup that says “ask our AI what to buy” is not automatically interesting anymore. ChatGPT, Google, Perplexity, Amazon, retailers, and browser extensions can all add conversational shopping features. The standalone startups that survive need a more specific advantage.
The evidence is stronger in narrower wedges. Phia has price and resale confidence. Onton has visual decision-making. Lily AI has product attributes. Constructor has enterprise discovery. ShopMy has creator-led commerce proof. Tilt has seller-side AI tooling. Cernel has agentic-commerce data infrastructure.
The weaker zone is the middle: broad shopping assistants with no disclosed user base, no merchant network, no conversion metric, and no proprietary data loop. Those companies may still become interesting, but today we should not crown them just because they use the right AI vocabulary.
Which AI shopping startups are the top ones overall right now?
The top AI shopping startups right now are Phia, ShopMy, Constructor, Gensmo, Daydream, Onton, Lily AI, Cernel, ShopOS, Alta, and Tilt.
Phia is the top emerging consumer AI shopping startup. It has the clearest mix of recent user growth, partner coverage, funding, and a product wedge people understand quickly. If we had to pick one pure AI shopping app that feels hottest today, it would be Phia.
ShopMy is the strongest commerce company around the AI shopping market. It is not the cleanest “AI assistant” label, but it has something most AI shopping startups still lack: serious monetization proof, creator supply, and transaction flow. These days, that is more convincing than another assistant demo.
Constructor is the top established infrastructure startup. It is less new, but retailers already need what it sells. In a market full of agentic promises, that makes it more durable.
Gensmo is the biggest high-upside AI fashion-agent bet. The funding is large enough to take seriously, but it still needs more public traction evidence to rank above Phia.
Daydream is a credible fashion discovery platform because of team, capital, and catalog breadth. The key question is whether it can create a new consumer habit.
Onton is the strongest emerging visual-shopping name because it has usage momentum and a clear starting wedge in home and furniture.
Lily AI is one of the most important infrastructure startups because product attributes are becoming the machine-readable layer of shopping.
Cernel is the freshest agentic-commerce infrastructure name. It is early, but its timing is excellent.
ShopOS is the commerce-operations startup to watch because it attacks the brand workflow side of AI shopping.
Alta is one of the better personal styling startups because it starts from closet memory rather than generic search.
Tilt is the live-shopping AI startup to watch because it uses AI to help sellers operate faster, not just to make buyers chat with a bot.
So, at the end, AI shopping does not have one winner yet. It has three leading zones. Phia leads the new consumer-confidence layer. ShopMy leads trust and creator-led commerce. Constructor, Lily AI, Cernel, and ShopOS are building the merchant infrastructure that agents and retailers will need. The strongest startups are the ones with proof in one of those zones, not the ones with the broadest “AI shopping assistant” pitch.
| Category | Startups selected and why |
|---|---|
| Consumer AI shopping apps | Phia leads because it combines fresh user growth, brand coverage, and a clear price-confidence wedge; Onton follows on active-shopper momentum; Daydream and Gensmo are bigger platform bets; Alta has the sharpest closet-memory angle |
| Funding versus proof | Phia has the best funding-to-usage balance; Gensmo has the largest early-stage AI fashion bet; ShopMy has the strongest commercial proof; Constructor is the mature B2B benchmark; ShopOS is the emerging commerce-ops bet |
| Fashion AI shopping | Phia leads on price and resale confidence; Daydream leads on catalog breadth; Gensmo leads on AI stylist ambition; Alta leads on personal wardrobe data; shoppin’ and BNTO are earlier but culturally well-positioned |
| Shopping confidence | Phia leads on purchase confidence; Onton leads on visual decision-making; Zellor is early but has a concrete conversion-lift signal |
| Merchant infrastructure | Constructor leads enterprise product discovery; Lily AI leads product-attribute infrastructure; Cernel is the freshest agentic-commerce data startup; ShopOS targets ecommerce operations; Athos shows consolidation pressure |
| Visual search and try-on | Gensmo has the biggest technical ambition; Alta has the best personal data loop; Onton has stronger visual-decision usage proof; shoppin’ and BNTO are smaller but aligned with social and circular shopping behavior |
| Social commerce | ShopMy is far ahead on creator scale and monetization; Tilt is strong in live shopping; Phia has cultural distribution; BNTO is an early styling-commerce community bet |
| Agentic checkout | Perplexity sets the reference point; Cernel builds the data layer; Constructor protects retailer discovery; ShopOS helps brands adapt operations; Phia can sit before checkout as a neutral confidence layer |
| Weak categories | Broad AI shopping chatbots with no usage, merchant network, conversion proof, or proprietary data loop still look too vague to rank seriously |
| Overall top startups | Phia, ShopMy, Constructor, Gensmo, Daydream, Onton, Lily AI, Cernel, ShopOS, Alta, and Tilt stand out after comparing recent usage, funding, merchant proof, infrastructure relevance, and workflow depth |
If you want more recent data on this point, please see our latest AI shopping market report.
This chart, featured in our AI shopping market deck, compares the main business model options for AI shopping assistants
OUR METHODOLOGY
This analysis tests which AI shopping startups look strongest now based on the evidence available today. We compare the companies across consumer usage, merchant coverage, funding versus proof, transaction flow, product specificity, infrastructure relevance, visual shopping strength, social-commerce depth, and agentic-checkout positioning.
The AI shopping market is still early, uneven, and easy to misread if treated as one broad ranking. To make the answer clearer, we broke the category into the dimensions that actually shape competitive strength: consumer usage, funding versus proof, shopping confidence, fashion discovery, merchant infrastructure, visual search, social commerce, and agentic checkout.
For each dimension, we looked at recent signals rather than relying on market intuition or generic AI-shopping narratives. We prioritized disclosed user traction, merchant coverage, revenue or transaction proof, conversion impact, catalog breadth, funding timing, product specificity, and infrastructure relevance.
The final ranking reflects that aggregation. We gave more weight to startups that showed proof inside a clear wedge, whether that was price confidence, creator-led commerce, product data, visual decision-making, live selling, or agent-ready infrastructure.
When we rank Phia highly, we are treating its reported users, brand partners, funding, celebrity distribution, and price-comparison use case as stronger current proof than a launch announcement or a broad AI-shopping pitch. The point is not that Phia has already won the whole market, but that its consumer-confidence wedge is unusually clear.
When we rank ShopMy highly, we are widening the lens beyond pure AI assistants. Its creator scale, revenue growth, profitability, valuation, and transaction flow make it one of the strongest commerce companies around the AI shopping market, even if it is not the cleanest “AI shopping app” label.
For infrastructure companies, we gave extra weight to businesses that solve problems retailers already understand: product discovery, product attributes, merchant data, ecommerce operations, and agent-ready catalog infrastructure. That is why Constructor, Lily AI, Cernel, ShopOS, and Athos Commerce appear in the analysis alongside consumer-facing apps.
For fashion and visual-shopping startups, we separated ambition from visible proof. Gensmo, Daydream, Alta, shoppin’, BNTO, and Onton all matter, but they are not interchangeable: some are stronger on funding, some on catalog depth, some on personal data, some on visual decision-making, and some on cultural discovery behavior.
We prioritized sources that added specific, checkable information: user counts, monthly active shoppers, brand partners, funding rounds, valuation, catalog size, product attributes, conversion lift, live-commerce tools, agentic-commerce infrastructure, and checkout features. We excluded commentary that repeated the AI-shopping theme without adding evidence.
Key sources used for this analysis include: Vogue on Phia’s users, brand partners, funding, celebrity investor base, and positioning, People on Phia’s $35.5 million round and product features, TechCrunch on Onton’s growth and expansion beyond furniture, PR Newswire on Daydream’s launch, product catalog, brands, and partners, Vogue on Daydream’s fashion discovery positioning, PR Newswire on Gensmo’s $60 million-plus funding, TechCrunch on Alta’s $11 million seed round and wardrobe product, Vogue on Alta’s product and founder context, ShopMy on its valuation announcement, PR Newswire on ShopMy’s $70 million raise and $1.5 billion valuation, PR Newswire on Constructor’s Series B and $550 million valuation, Lily AI on product attributes and training data, Cernel on its €4 million seed funding, EU-Startups on Cernel’s agentic-commerce infrastructure thesis, Think Business on Zellor’s funding and AI video shopping assistant, Silicon Republic on Zellor’s €850,000 round, Athos Commerce on the Klevu and Searchspring combination, Business Wire on Athos Commerce’s AI-backed ecommerce platform, Tilt on its $26 million funding round, Business Insider on Tilt’s AI seller tools, TechCrunch on Perplexity’s shopping launch, The Verge on Perplexity’s Buy with Pro checkout flow, and The Times of India on ShopOS’s $20 million funding and ecommerce operations focus.
This chart, featured in our AI shopping market deck, shows how market revenue is split across customer segments in the AI shopping market
Related blog posts
- How strong is fundraising in the AI shopping market right now?
- The startups that have raised the most funding in the AI shopping market
