Case study · Ciepło24.pl
An AI duo: no more "no results" and no more queue of the same questions
How a heating store replaced a search engine that lost customers and an overloaded service team with a pair of AI assistants: ecSearch + ecChat.
Client and context
Ciepło24.pl (technika-grzewcza-sklep.pl) - an online store and wholesaler of heating technology: boilers, controls, chimney systems, water heaters and fittings from over 30 brands. The store runs on an e-commerce platform we built and continue to develop - and its customers, from consumers to installers, ask about technical products, each in their own way.
Challenge
Customers searched for products by symptoms ("boiler not heating"), colloquial names and with typos - the standard search engine replied "no results". The service team answered the same questions dozens of times a day.
Solution
We deployed an AI duo: ecSearch (semantic search with autocorrect and suggestions) and ecChat (an assistant advising 24/7 and checking availability and order statuses).
How it works
Two assistants, one customer flow
ecSearch catches the customer at the moment of searching, ecChat - at the moment of doubt. Together they guide the customer from the first query all the way to placing the order.
A search engine that understands intent
- Semantic search instead of matching letters
- Autocorrect for typos and word forms ("monobloc heat pumpp")
- Understands symptoms: "boiler not heating" leads to the right category
- Suggestions after 3 characters, before the customer makes a mistake
An adviser that never sleeps
- Advises on product selection in the heating sector
- Checks availability and order status in real time
- Answers 24/7 - even when the service team is off duty
- Hands harder cases over to an agent with full context
Results
The search engine sells, the assistant answers
Deployment process
From audit to production in 2-4 weeks
No store engine replacement - we connected the AI duo to the existing platform through integration.
Analysis and data
Audit of no-result queries and the most frequent customer service questions, gathering product knowledge.
Integration
ecSearch on the catalogue, ecChat wired to stock levels and order statuses.
Training Cieplik
Teaching the assistant the specifics of the heating sector and the brand's tone, testing on real questions.
Launch and tuning
Going live on real traffic and ongoing monitoring of answer quality.
Deployment stack
Let's talk
Want similar results?
We will show you the ecChat + ecSearch duo on examples from your industry and estimate the potential impact. 30 minutes, no obligation.