Conversational analytics over your data
Don’t ask passers-by about your revenue.
AI chat can answer almost anything. But it can only answer questions about your numbers correctly when it knows your data. We build a data warehouse with a semantic layer — turning AI chat into your analyst.
Online, non-binding and free.
What was our revenue in September?
Illustrative data of a fictitious online store
Three ways to ask about revenue
Asking an AI about your sales without context is like a CEO going out on the street and asking passers-by instead of asking their own data analyst.
01
You ask a passerby
AI chat without your data. It readily answers, but knows nothing about your business. You get general advice or an honest “I don’t know”.
What was our revenue in September?
I don’t have access to your company’s data, so I don’t know your revenue. In general, September tends to be…
Illustrative data of a fictitious online store
02
You give a passer-by a box of invoices
AI working with an export from your online store or accounting system. It answers confidently with a specific number. But it chose its own definition of revenue: with or without VAT? Before or after returns? Based on Shoptet orders or Pohoda invoices? Including unpaid invoices or paid only?
This step hurts. And people do it every day.
What was our revenue in September?
Your revenue in September was 4,850,000 CZK.
- With VAT?
- With shipping?
- Before returns?
- Orders or invoices?
- Paid only?
Illustrative data of a fictitious online store
03
You ask your own analyst
AI working with a data warehouse and a semantic layer. It gives you the number you agreed on and explains how it was calculated: the definition, sources and period.
What was our revenue in September?
Revenue in September: 3,527,000 CZK.
- Definition: excluding VAT · excluding shipping · after returns · by invoice date · including unpaid invoices
- Sources: Pohoda, Shoptet
- Period: 1–30 September
Illustrative data of a fictitious online store
How many versions of the truth does your revenue have?
One question, five decisions, 32 different numbers. Switch and watch the 'revenue' change.
What was our revenue in September?
4,850,000 CZK
Difference between highest and lowest version: 1,394,000 CZK
The semantic layer makes this decision once, permanently and for all. Each AI then answers the same.
Semantic layer →Show all 32 variants
| VAT | Shipping | Returns | Source | Payment | Revenue |
|---|---|---|---|---|---|
| with VAT | included | before deduction | orders (Shoptet) | including unpaid | 4,850,000 CZK |
| with VAT | included | before deduction | orders (Shoptet) | paid only | 4,753,000 CZK |
| with VAT | included | before deduction | invoices (Pohoda) | including unpaid | 4,705,000 CZK |
| with VAT | included | before deduction | invoices (Pohoda) | paid only | 4,610,000 CZK |
| with VAT | included | after deduction | orders (Shoptet) | including unpaid | 4,559,000 CZK |
| with VAT | included | after deduction | orders (Shoptet) | paid only | 4,468,000 CZK |
| with VAT | included | after deduction | invoices (Pohoda) | including unpaid | 4,422,000 CZK |
| with VAT | included | after deduction | invoices (Pohoda) | paid only | 4,334,000 CZK |
| with VAT | excluded | before deduction | orders (Shoptet) | including unpaid | 4,680,000 CZK |
| with VAT | excluded | before deduction | orders (Shoptet) | paid only | 4,587,000 CZK |
| with VAT | excluded | before deduction | invoices (Pohoda) | including unpaid | 4,540,000 CZK |
| with VAT | excluded | before deduction | invoices (Pohoda) | paid only | 4,449,000 CZK |
| with VAT | excluded | after deduction | orders (Shoptet) | including unpaid | 4,399,000 CZK |
| with VAT | excluded | after deduction | orders (Shoptet) | paid only | 4,311,000 CZK |
| with VAT | excluded | after deduction | invoices (Pohoda) | including unpaid | 4,267,000 CZK |
| with VAT | excluded | after deduction | invoices (Pohoda) | paid only | 4,182,000 CZK |
| without VAT | included | before deduction | orders (Shoptet) | including unpaid | 4,008,000 CZK |
| without VAT | included | before deduction | orders (Shoptet) | paid only | 3,928,000 CZK |
| without VAT | included | before deduction | invoices (Pohoda) | including unpaid | 3,888,000 CZK |
| without VAT | included | before deduction | invoices (Pohoda) | paid only | 3,810,000 CZK |
| without VAT | included | after deduction | orders (Shoptet) | including unpaid | 3,768,000 CZK |
| without VAT | included | after deduction | orders (Shoptet) | paid only | 3,692,000 CZK |
| without VAT | included | after deduction | invoices (Pohoda) | including unpaid | 3,655,000 CZK |
| without VAT | included | after deduction | invoices (Pohoda) | paid only | 3,582,000 CZK |
| without VAT | excluded | before deduction | orders (Shoptet) | including unpaid | 3,868,000 CZK |
| without VAT | excluded | before deduction | orders (Shoptet) | paid only | 3,791,000 CZK |
| without VAT | excluded | before deduction | invoices (Pohoda) | including unpaid | 3,752,000 CZK |
| without VAT | excluded | before deduction | invoices (Pohoda) | paid only | 3,677,000 CZK |
| without VAT | excluded | after deduction | orders (Shoptet) | including unpaid | 3,636,000 CZK |
| without VAT | excluded | after deduction | orders (Shoptet) | paid only | 3,563,000 CZK |
| without VAT | excluded | after deduction | invoices (Pohoda) | including unpaid | 3,527,000 CZK |
| without VAT | excluded | after deduction | invoices (Pohoda) | paid only | 3,456,000 CZK |
Illustrative data of a fictitious online store.
What you may ask
Conversational analytics means asking about your own data in ordinary speech — in the AI chat you already use.
What the AI needs to know to answer correctly
- How do you calculate your margin?
- Purchase prices from accounting (Pohoda, Money S4, Abra)
- Returns and cancellations from the online store
- Which category belongs to which product
This is exactly what the semantic layer contains: the analyst's knowledge, written down once.
Conversation analytics →How it works
Resources
- Shoptet
- Pohoda
- Money S4
- Abra
- GA4
- Google Ads
- Meta
- Sklik
AI chat
Click to change the tool
You replace the tool. The context remains.
- 01
We connect your systems
Shoptet, Pohoda, Money S4, Abra, GA4, Google Ads, Meta, Sklik and others. Data is refreshed automatically.
- 02
We will build a cloud data warehouse
All data in one place and with history. Today we build on Google Cloud and BigQuery, but the value is not in the technology, it is in the data model.
- 03
We will propose a semantic layer
Uniform definitions of metrics: what is sales, margin or active customer. Names, descriptions, calculations, and relationships between data.
- 04
You connect the AI chat you already use
ChatGPT, Claude, Gemini, Copilot, or whatever comes next.
Before we hand over the warehouse to you, we test the AI on a set of real questions with known answers. You will see the results.
No one knows what AI will be the best in a year. But each will need your context.
- Passersby will respond equally to you and your competition. Only you have your own analyst.
- Knowledge stays with the company, not the tool. Every new AI tool becomes your analyst right away.
- It takes months to train an analyst. Context too — whoever starts now has a head start.
They invested in context
Connect new AI agents to reliable data and scripted processes now.
They only invested in AI tools
Tools respond unreliably and data cleaning begins under pressure.
They waited
They start from scratch, behind the competition.
Investing in context wins in all three scenarios.
Why start now?
The history you don't capture is gone.
Daily inventory levels or price lists are usually overwritten by source systems. Those who don't save them don't have them.
Definitions must be agreed upon.
What is a margin or an active customer is a management decision, not a technical detail. It takes time — and it's better to get it done before the AI starts responding.
Your people are already using AI.
Exports to AI chats are already uploading today. Without uniform definitions and without control over where the data goes.
Agents raise the stakes.
A chat that guesses will give the wrong answer. An agent who guesses makes a wrong move.
More than numbers: the corporate brain
A good analyst knows not only the numbers, but also the reasons behind them. The corporate brain (second brain) is a structured knowledge base — strategy, processes, clients, decisions from meetings — that is read by both humans and AI. It is the natural next step after the data warehouse. And we drive Solvik on it ourselves.
Why AIcontext by Solvika
Behind AIcontext is Solvika — business intelligence for dozens of companies in the Czech Republic and Slovakia.
We know the systems you are running on
We connect Shoptet, Pohoda, Money S4, Abra, GA4, Google Ads, Meta and Sklik. We know what their data means — and that's what the semantic layer consists of.
We test before we deliver
The AI must correctly answer a set of real questions with known answers. Only then will we hand over the warehouse to you.
We use it ourselves
We run Solvika over our own corporate brain, which is read by humans and AI agents. We sell what we use ourselves.
Not tied to a single AI vendor
The value is in the data model and semantic layer, not the specific tool. You connect what suits you.
Business Intelligence in Qlik Sense
We specialize in the design, implementation and development of Qlik Sense dashboards.
We build reports in business language — so that the owner, marketing manager and financial director can understand them, not just IT.
Free consultationCustom dashboardsImplementation of Qlik Sense
Analysis of data and processes, design of dashboards, connection of source systems, deployment and training of the team.
Data Integration (ETL)
Connection of online store, ERP, accounting, CRM and marketing tools into one data model.
Custom dashboards
Sales, margins, customers, retention, acquisition, CLV, assortment, inventory, marketing campaigns, regional sales.
Administration and development
Ongoing maintenance, expanding reports, support, training.
Reference
Verified by dozens of companies
See what our clients say about the cooperation.
Fundamental acceleration of analysis
„Nástroj Qlik od Solvika mi zásadně změnil způsob, jak analyzuji data e-shopu. Výrazně mi zrychlil práci s čísly — akvizice, retence, tržby, produkty. Jsem nadšený jako už dlouho z ničeho.“
Jiří Doležal
mediskont.cz
Cooperation beyond expectations
„Skvělý analytický nástroj, který nám výrazně zpřehlednil a zefektivnil řízení. Oceňujeme úpravy na míru přesně podle našich potřeb. Spolupráce předčila naše očekávání po všech stránkách.“
Marcel K.
fujifoto.cz
A new look at data
„Vďaka riešeniam v Qlik Sense máme konečne všetky dôležité dáta prehľadne a zrozumiteľne. Získali sme nový pohľad na predaje, marže aj správanie zákazníkov.“
Milan Kozický
medaren.sk
Profit maximization
„Kompletní přehled zákazníků, objednávek, obratů a tržeb. Sledování krajů, obchodních zástupců — nepřeberné množství možností pro maximalizaci zisku. Doporučuji všem.“
Vít Kolčaba
czechpodsb2b.cz
More effective campaigns
„Pomáha nám optimalizovať a správne cieliť reklamné kampane. Máme podrobný prehľad o efektivite kampaní a o vývoji predaja. Práca s Qlik Sense je jednoduchá, na pár klikov máte relevantné informácie.“
PLZR
plzr.sk
Faster decision making
„Spolupráce s firmou Solvika a implementace Qlik Sense byla skvělým rozhodnutím. Nástroj nám ušetřil čas a zefektivnil rozhodování. Podpora je vždy rychlá a profesionální.“
Tomáš Jurka
bushcraftshop.cz
Frequently asked questions
ChatGPT tells me it doesn't know our sales. He's not making anything up, so where's the problem?
You are right, an honest passerby will say 'I don't know'. The problem starts when you give him a box of invoices — upload an export to the chat. Then it answers confidently, but according to the definition it has come up with. And this happens every day in companies.
Won't AI soon be so smart that it can handle our data by itself?
Even the economics professor you meet on the street doesn't know your sales. Smarter models think better, but intelligence is not information: no model knows that your margin does not include transport unless it is written somewhere.
Isn't it worth waiting for the AI market to calm down?
It won't calm down — and you don't need it. Context is the part of preparation that doesn't depend on which tool wins. You save nothing by waiting, you just put off what takes the longest.
We have our own analyst. What is it to us?
The semantic layer is your analyst's knowledge, written once. Everyone gets answers according to his definitions — and it stops answering the same question for the fifth time.
We already have BI and dashboards. Do we need it?
Then you are half done. Dashboards answer pre-given questions, AI chats and those that no one expected — but only above the semantic layer. Your BI model is the starting point.
Aren't we too small for a data warehouse?
Cloud data warehouses are paid according to actual usage, not flat-rate licenses. The main work is in the data model and semantic layer — and we have that ready for the systems you use. How much it would cost you, we will discuss at the consultation.
Is our data safe when AI can access it?
More secure than today when people put exports in chats: single source, clearly set access rights and AI tools connected with explicit permission. The warehouse can also run in your own cloud project — we'll agree. We will discuss specific settings during the consultation.
Which AI tool should we choose?
The one your team is already using. With the finished context, the choice of tool is reversible.
How do we know the AI is answering correctly?
Before handover, we test the AI on a set of real business questions with known answers and tune the semantic layer until it passes the test. You will see the results.
How much does it cost and how long does it take?
It depends on how many systems we connect and how complex your definitions are. We will tell you specifically during the free consultation.
Make AI chat your analyst
Free and non-binding online consultation. We'll go through your systems and show you where to start.
















