The most expensive part of any AI project isn't the model. It's the data you feed it. For large retailers managing hundreds of thousands of product SKUs, that data is often a mess of inconsistent spreadsheets, missing attributes, and manual entry errors. Merchkit, a Vancouver and Toronto-based startup, is betting that cleaning up this foundational layer is a more valuable wedge than selling another generative AI content tool.
Founded in 2022, the company has built software that automates the ingestion, enrichment, and publishing of retail product catalogs. Its system takes raw, incomplete product data, cleans and structures it, then uses AI to generate optimized titles, descriptions, and attributes before pushing the finalized listings to channels like Shopify, Amazon, and Walmart [BetaKit, February 2025]. Early customers, including home improvement giant Lowe's, have reported efficiency improvements of up to 90% on routine catalog tasks [Techcouver.com, March 2025].
The Wedge: Data First, AI Second
Merchkit's approach is deliberately infrastructural. Instead of starting with a flashy AI content generator, the product focuses on the unglamorous work of data normalization. The premise is that AI can only deliver reliable, scalable value if the underlying product information is consistent and machine-readable. The company's platform is designed to handle catalogs ranging from 5,000 industrial components to 500,000 SKUs, transforming them into what it calls "AI-ready feeds" [Merchkit | B2BEA, retrieved 2026].
This focus on the data pipeline shapes the product's reported outcomes. The cited efficiency gains for early customers stem from automating manual data entry and cross-referencing, not just from faster content creation. For retailers in furniture, appliances, and apparel,segments where Merchkit is reportedly building a presence,product attributes like dimensions, materials, and compliance data are critical and often poorly documented [BetaKit, February 2025]. Automating this cleanup creates a defensible cost-saving case before any AI-generated copy is ever published.
A Founding Team Built for Scale
The company is led by a trio of founders with complementary backgrounds in scaling software businesses and product-led growth.
| Founder | Role | Key Prior Experience |
|---|---|---|
| Bijan Vaez | CEO | Co-founder & former CTO of EventMobi; former CTO of Browze [BetaKit, February 2025]. |
| Rob Hayes | Founding Team | Former Head of Product at Voiceflow and Ada; former Partner at First Round Capital [BetaKit, February 2025][Rob Hayes, Seedlist.com, retrieved 2026]. |
| Vasil Damyanov | Founding Engineer | Former CTO of ContentFly [BetaKit, February 2025]. |
Vaez brings experience as a three-time founder who helped scale EventMobi to millions of users [Artificial Insights: How to Build an AI Business podcast, retrieved 2026]. Hayes adds deep product craft from the conversational AI space and a long-term perspective from over a decade in venture capital. This blend suggests a team oriented toward both enterprise product execution and the strategic motions required to build a venture-scale company in a competitive retail tech landscape.
Traction and the Road from Pre-Seed
In late 2024, Merchkit closed a $1.4 million (USD) pre-seed round. The financing was led by Afore Capital and Antler Canada, with participation from undisclosed strategic angels in retail and technology [BetaKit, February 2025]. The capital is earmarked for expanding the team and accelerating product development.
The company's early beachhead is significant. Landing Lowe's as a customer provides a referenceable enterprise case study in a complex, high-SKU environment. Merchkit reports having "a handful" of other enterprise contracts and is actively developing its presence in the furniture, appliance, and apparel verticals [BetaKit, February 2025]. The initial value proposition appears to be resonating with businesses where catalog management is a known, painful, and labor-intensive cost center.
The Technical Breakdown: Where the Rubber Meets the Road
From an infrastructure perspective, Merchkit's system has to solve several hard problems concurrently. It must ingest data from a myriad of source formats, apply validation and deduplication logic, and then orchestrate a pipeline where AI models augment the data without introducing hallucinations or brand-voice inconsistencies. The 90% efficiency claim points to a high degree of automation in data matching and attribute population, tasks that typically require human review.
The real technical risk isn't the AI layer, but the data orchestration underneath it. At scale, with half a million SKUs, pipeline failures, integration breakages with channel APIs, or slow feedback loops for model correction could erode the very efficiency gains the product promises. The system's resilience will be tested as catalog update frequencies increase and as it expands from a "handful" of enterprise customers to dozens.
Navigating a Crowded and Evolving Landscape
The retail tech space is densely populated, though Merchkit's specific focus on catalog data automation provides a clear point of differentiation. The company is not directly competing with broad e-commerce platforms or standalone AI copywriting tools. Instead, it sits upstream, ensuring the fuel for those tools is high-quality. However, this positioning creates its own set of challenges.
- Platform expansion. Large retail platforms like Shopify or Amazon could decide to build or acquire similar native catalog management capabilities, potentially squeezing a point-solution vendor.
- Economic sensitivity. Retailers facing margin pressure may cut spending on software perceived as a "nice-to-have" efficiency tool, rather than a direct revenue driver.
- Customization burden. Enterprise retailers often have unique data schemas and legacy systems. The cost of implementation and customization could slow sales cycles and dilute gross margins.
The company's most plausible answer to these pressures is depth. By going deep on the data problems of specific, high-SKU verticals, Merchkit can build a level of domain-specific understanding and workflow automation that a general-purpose platform would struggle to replicate quickly. Its success will depend on proving that its specialized tool delivers a return on investment that is both rapid and substantial enough to justify the ongoing cost.
The next twelve months will be about proving repeatability. The Lowe's win is a strong start, but the key metric to watch is the expansion of that initial footprint into other large retailers within its target verticals. If Merchkit can standardize its implementation playbook and demonstrate consistent efficiency gains, it will have a compelling story for a larger seed round to fuel a direct sales push. The bet is that in the age of AI, the most valuable software might just be the plumbing that makes everything else work.
Sources
- [BetaKit, February 2025] Former EventMobi, Voiceflow leaders secure $2 million CAD to tackle retail merchandising with Merchkit | https://betakit.com/former-eventmobi-voiceflow-leaders-secure-2-million-cad-to-tackle-retail-merchandising-with-merchkit/
- [Techcouver.com, March 2025] MerchKit Reveals $2M Investment to Power Automated Merchandising Tech | https://techcouver.com/merchkit-reveals-2m-investment-to-power-automated-merchandising-tech/
- [Merchkit | B2BEA, retrieved 2026] Merchkit company profile | https://b2bea.com/company/merchkit
- [Artificial Insights: How to Build an AI Business podcast, retrieved 2026] The AI Strategy That Makes Business More Human with Bijan Vaez, CEO @ Merchkit | https://rss.com/podcasts/manaryhaus/1940948/
- [Rob Hayes, Seedlist.com, retrieved 2026] Rob Hayes profile | https://seedlist.com/people/rob-hayes