Pigu.lt integrates AI at the core of e-commerce
Leading online retailer faced a dilemma: refining product descriptions and QAing 4.5 million products from a 6.5M catalog. Here’s how we did it with AI in 6 weeks.
99.8%
Cost reduction on AI product processing
4.5M
Product descriptions refined with AI in 1.5 months
By Mia Lysikova
Client at a glance
Meet Pigu.lt
Pigu.lt is the leading e-commerce center in the Baltics, with a massive catalog of 6.5 million products across multiple languages, categories, and markets.
They manage 4.5 million products needing enrichment to meet quality standards for traffic and sales, and a constant flow of new products that require quality assurance: translation, sensitive content filtering, pricing, and more. Manual QA became too costly and lengthy.
Before NexEco AI
Scaling leading online retailer in Northern Europe
Pigu.lt faced a massive task:
Manually enriching 4.5M products
The process was slow, expensive, and created two-week backlogs. They spent $350K on translations alone.
Quality issues
Quality issues kept appearing with pricing, filtering inappropriate content, and product information. After working with consultants for 1.5 years with no results, they needed a better solution.
“Our content team needed to optimize 4 million products in our catalog – that’s a lot. And to do it by hand would take 3 years and cost us millions of euros. That’s extremely expensive. With NexEco AI, optimizing one item is 99.8% cheaper than doing it manually.”
Vytautas Romeika
CMO @ Pigu.lt
With NexEco AI
Integrating AI into the core of e-commerce
Massive cost reduction
Previously, the Pigu.lt team spent huge budgets to optimize each product description.
By using AI Gateway to automatically generate and improve product descriptions, product enrichment costs dropped by 99.8%.
Fast AI processing
Production delays for product descriptions and search optimization extended to 2+ weeks, impacting time-to-market.
From:
3 years
To manually process all QA queries
With NexEco AI, Pigu.lt completed catalog enrichment in 1.5 months versus an estimated 3 years for manual processes of all products.
To:
1.5 months
Took for AI to process all products with NexEco AI
Automatic image filtering
Certain inappropriate photos cannot be uploaded to the platform according to regulations. Manual QA takes a long time.
Instead of a dedicated team of human QA agents, Pigu.lt sends a request through the AI Gateway — if the AI model identifies that the image includes prohibited symbols or other imagery, it won’t allow the product to be uploaded.
Product translation at scale
Pigu.lt spent $350K on external translation agencies when sellers needed product information in additional languages.
From:
$350K
spent on external translation agencies
With NexEco AI, Pigu.lt can translate product descriptions for approximately $30K, reducing translation costs by over 90%.
To:
$30K
AI-generated product translations
Automated intelligence for sensitive categories
Pigu.lt struggled to manually create ingredient lists for beauty products and similar items to meet regulatory requirements, creating a significant operational bottleneck.
Using NexEco AI, they automatically retrieve ingredient information based on product barcodes and autofill the required fields, ensuring regulatory compliance with minimal effort.
Instant market pricing
Pigu.lt struggled to manually create ingredient lists for beauty products and similar items to meet regulatory requirements, creating a significant operational bottleneck.
Using NexEco AI Gateway, the platform automatically gathers market data to recommend optimal price ranges and implement safeguards that prevent sellers from abusing pricing.
Future
What’s next for Pigu.lt x NexEco AI
Pigu.lt is taking AI integration to the next level by turning it into a revenue stream. As they scale to 8 million requests, this price difference transforms AI from a cost-saving tool into a profitable business offering, creating value for all parties in their ecosystem.
