Skip to content
Sign in
SWC 2026 FinalistGrand Final: Main Stage

4sell

Edge-AI service-quality analytics and staff gamification that lift upsell revenue in cafés, retail stores and pharmacies.

Astana, Kazakhstan, AsiaAI & software, AI for upselling and service quality controlAzerbaijan (Baku)
Stage
Pre-seedAI at the core
Product
On the marketFirst customers
Funding raised
$400Kdisclosed
Founded
20251 years ago
Founders
2Roman, Konstantin
Global score
4.5Grade DRank 63 of 151

Overview

4sell.ai, a startup founded in Kazakhstan and resident at Astana Hub, helps retailers increase their upsell revenue through automated service quality control and team gamification. Winner of the Startup World Cup Azerbaijan regional final (Baku ID 2026, September 2026).

Traction

First customers

Solution deployed in retail outlets (retail, restaurants, pharmacies; logos Hani, SF, Expresso… on the website; a partner café earned $53 more in 5 hours). Target: 1,000 locations by end of 2026 and US sales by end of 2026 or early 2027; current number of customers not published.

Market

AI in retail

Physical retailers and restaurants lose revenue when staff fail to suggest add-ons or deliver consistent service, and managers lack visibility across sites. 4sell automates service-quality monitoring and gamifies upselling, monetising measurable revenue uplift per location. Cheap edge AI makes in-store analysis viable for SMEs, and labour costs make staff productivity a priority. The sale is location by location, so channel partners matter.

Addressable market

2025
$9.4B
2030
$23.6B
+20.3% a year

AI in retail: $9.4bn in 2025, +20.3%/year (Expert Market Research); Fortune BI estimates $12.4bn.

Customers

Retail stores, restaurants/cafés and pharmacies (logos Hani, SF, Expresso on the website).

Who signs the cheque

Owners and operations managers of multi-site retailers and F&B chains.

Geographies

  • Kazakhstan
  • United States (planned)

Drivers

  • Labour costs and staff turnover
  • Cheap edge AI hardware
  • Pressure to grow average ticket size

Headwinds

  • Privacy and employee-monitoring regulation
  • Competition from retail analytics tools
  • Anecdotal evidence of impact

The AI in retail market is valued at $9.4bn in 2025 (+20.3%/year, $40.7bn in 2033) according to Expert Market Research; Fortune BI estimates it at $12.4bn.

Scalability 7/10. AI software easily replicable in retail; constraint: strong competition and store-by-store sales.

Market source: marketresearch.com

Technology

Category

Retail AI SaaS with edge inference

AI at the coreMachine learning / predictiveLanguage / voice processing

Maturity

  1. R&D
  2. Prototype
  3. Pilot
  4. On the market
  5. Scaling

in market

How it works

AI analyses service interactions at the point of sale to score quality and coach upselling, with gamification for staff. An edge model is being developed to process data on-site; member of NVIDIA Inception.

Moat

Weak to moderate: data on upsell behaviour per site could become a moat, but the concept is replicable and evidence of uplift is anecdotal.

Differentiation

Combines quality control, upsell coaching and gamification in one tool versus pure analytics dashboards or mystery shopping.

Integration points

  • POS systems
  • CCTV/audio devices
  • Customer review and reputation platforms
  • Revenue/margin analytics agents

Depends on

  • Privacy compliance for audio/video analysis
  • POS data access
  • Edge hardware supply

Use of AI

Automated AI analysis (including edge AI) of point-of-sale service quality to drive upselling.

Product in detail

Automated AI analysis of service quality, staff motivation and gamified upselling; development of an edge AI model that processes data directly at the point of sale. Member of the NVIDIA Inception program (2026).

Business

Model
SaaS per location (pricing not published).
Revenue logic
Value tied to upsell revenue gains per site (e.g. a partner café earned $53 more in 5 hours).
Go-to-market
Direct local sales from Astana Hub; target 1,000 locations by end of 2026 and US sales by end of 2026 or early 2027.

Next milestones

  1. 1,000 locations live
  2. A controlled uplift study with a chain
  3. First paying US customer

Key risks

  • Weak proof of impact
  • Small home market
  • Small capital base ($400,000)

Opportunities

4 found in the network

Concrete moves for 4sell inside the network: who to sell to, sell through, build with or raise from, each with the reasoning, the evidence and the first step.

  • Distribution 1
  • Market entry 1
  • Co-development 1
  • Investors 1

Distribution1

A channel or customer base to reach together

Close the loop between online reviews and in-store service

Custplace manages reviews and reputation for franchise networks and key accounts in France and beyond. Bundling 4sell's in-store service scoring would let networks link online ratings to on-site behaviour, and gives 4sell a channel into large franchise networks.

25%confidence

Evidence from the records

  • “automated service quality control and team gamification”
  • “SaaS used by franchise networks and key accounts (Verisure in 18 countries, Direct Assurance, Optical Center, Darty, EDF, Stych)”
If it works
A joint CX offer for franchise networks and a first Western European channel for 4sell.
First step
Propose a joint pilot to one Custplace franchise customer with stores where service quality is measurable.

Market entry1

A local partner where they want to go

Access Latin American restaurant chains

Pulster serves 49 brands in Latin America, including quick-service chains such as McDonald's, KFC and Burger King, with location analytics. Adding store-level service and upsell metrics to Pulster's competitive analysis gives 4sell a channel into these chains.

15%confidence

Evidence from the records

  • “Solution deployed in retail outlets (retail, restaurants, pharmacies”
  • “49 brands in Latin America including McDonald's, KFC, Burger King, Subway, Walmart, OXXO”
If it works
Pilot introductions to Latin American QSR franchisees.
First step
Share a one-page joint value proposition with Pulster's partnerships lead.

Co-development1

An integration or joint product

Combine in-store behaviour data with margin-optimisation agents

Pardon AI builds agents that turn business data into revenue and margin decisions; 4sell captures what happens at the counter. An integration would let operators see which staff behaviours drive margin and act automatically, a stronger product for both, particularly in food service.

20%confidence

Evidence from the records

  • “Automated AI analysis of service quality, staff motivation and gamified upselling”
  • “AI agents that analyze business data to optimize revenue and margins.”
If it works
A combined product for F&B chains and shared US go-to-market.
First step
Run a technical call to define an API exchange of 4sell service scores into Pardon's data model.

Investors1

A fund whose thesis fits

US market-entry platform for retail AI

4sell plans US sales by end of 2026 or early 2027 and is raising a bridge. Plug and Play invests at Seed and Series A globally, covers Commerce & retail, and backed Yo-Kai Express in the network; its corporate retail programmes are a direct route to US pilots.

25%confidence

Evidence from the records

  • “US sales by end of 2026 or early 2027”
  • “Commerce & retail”
  • “Chikaranomoto Holdings (Ippudo), Japan Tobacco, Plug and Play”
If it works
Corporate retail pilots in the US and possibly a seed investment.
First step
Ask Yo-Kai Express for an introduction to its Plug and Play contact and apply to the retail programme.

Analyst reading of 21 Nov 2026, from the public record of the network. Method: analyst-ai.

Investors that fit

Kazakhstan's international IT technopark running pilots and programmes for regional startups, such as Hi Doctor's Smart City pilot.

Kazakh AI upselling for retail outlets with a deployed product: home-market startup, AI-core, pre-seed.

55%confidence

Evidence from the records

  • “AI for upselling and service quality control”
  • “Kazakhstan”
Intro path
Axiom Capital (4sell investor; deal sourced via Astana Hub media)
First step
Apply for Astana Hub residency and pilot programmes.

Investor with no public profile in our data; wrote a $110K pre-seed cheque into Kazakh smart-insole startup MiraiTech.

Kazakh AI upselling at pre-seed — same ecosystem and stage.

30%confidence

Evidence from the records

  • “pre-seed”
  • “AI for upselling and service quality control”
  • “Solution deployed in retail outlets (retail, restaurants, pharmacies”
Intro path
Founder-to-founder intro via MiraiTech (network portfolio company of JAS Ventures).
First step
Ask MiraiTech for an introduction.

Tokyo commerce-focused VC that joined Kauche's 2025 Series C in social commerce.

Kazakh AI upselling — retail AI.

25%confidence

Evidence from the records

  • “Commerce & retail”
  • “AI for upselling and service quality control”
  • “Solution deployed in retail outlets (retail, restaurants, pharmacies”
Intro path
Founder-to-founder intro via Kauche Inc. (network portfolio company of New Commerce Ventures).
First step
Low priority.

Investment analysis

4.5D

Grade D

Rank 63 of 151

Disruption4.0/10
Business model4.0/10
Market5.0/10
Risk(lower is better)6.0/10

Global score = mean of disruption, model, market and (11 − risk). Last updated 21 Nov 2026.

Local SaaS project to watch from a distance, not yet investable at scale.

Disruption

4/10

AI for service quality control and gamification for upselling: incremental improvement of existing tools.

Business model

4/10

Deployed in retail outlets with customer logos, but revenue not quantified and evidence anecdotal.

Market

5/10

TAM $9.36bn → base 5; scalability 7/10 (neutral); no international presence: Kazakhstan, US entry only planned → 5/10.

Risk

6/10

Lower is better

Medium-high risk: small amount raised ($0.4M) and competition from retail analytics tools.

Strengths

  • First active customers
  • Low acquisition cost in the local market

Risks

  • Anecdotal evidence of impact
  • Small home market

Funding

$400K, disclosed
  1. déc. 2025

    pre-seed ($4M valuation)$400,000

    MA7 Ventures, MOST Accelerator Fund, Axiom Capital

  2. 2025-2026

    grants and cloud credits (non-dilutive)more than $500,000

    major cloud providers and startup programs (not named)

  3. 2026

    bridge (under way, for edge AI and US market entry)n/a

    n/a

Amount used in the study. $400k pre-seed used; >$500k in grants and cloud credits (mix cannot be separated) in the upper bound; ongoing bridge excluded.

Note. ≈ $400,000 (Dec. 2025 round); more than KZT 205M in total according to the Kazakh government

The Tech and Astana Hub (30 June 2026) confirm admission to NVIDIA Inception, more than $500,000 in grants and cloud credits, and an open bridge round (amount not specified). Caplight confirms the 9 Dec. 2025 round: $400,000 at a $4M valuation. Azerbaijan regional (Baku ID 2026) unchanged.

Team

  • Roman Belopolsky

    Co-founder & CEO

    From Karaganda; co-founded 4sell.ai with school friends.

  • Konstantin Balzin

    Co-founder & CTO

    School friend from Karaganda, technical lead of 4sell.ai.

Sources

Regional final
Azerbaijan (Baku)
Identification confidence
high
Founding year: 2025 (low confidence)
"Founded 2025" according to a startup directory (PremierAlts); $400k raised in December 2025. No registry confirmation. premieralts.com
  1. 14sell.ai/
  2. 2finanznachrichten.de/nachrichten-2026-09/69680479-press-release-news-baku-id-202
  3. 3swacenews.com/2026/10/13/baku-id-2026-innovation-festival-concludes/
  4. 4the-tech.kz/kazahstanskij-ai-startap-4sell-ai-stal-uchastnikom-programm
  5. 5caplight.com/company/4sell
  6. 6astanahub.com/kk/article/karagandinskii-ai-startap-privliok-400-tysiach-o

From the open study Startup World Cup 2026 by TIBOK, based on public sources, verified in November 2026. Scores are an analyst's judgment on an explicit grid, not the jury's.

Wall

Wall

No posts yet