Project highlights
Task
Design an AI-powered system to automate data entry, unify datasets, and make bookstore analytics more accessible
Team
1 Product Designer
Duration
6 weeks
Scope
UX/UI design, AI-assisted workflows, automated visualization
About the product
Running an online bookstore means dealing with data everywhere: inventory records, product descriptions, sales dashboards, and customer details. But keeping this data consistent takes countless hours of manual work. CRMs and CMS platforms demand endless input, and dashboards often fall short of delivering real insights.
The goal was to design an AI data management tool for bookstore data: a system that could handle repetitive tasks like data entry and even generate descriptions and dashboards, making complex analytics accessible to non-technical staff.
Design focus
To achieve this, the platform had to:
-
reduce time spent on repetitive data entry across products and customer records;
-
give non-technical users a way to generate descriptions for thousands of titles without external tools;
-
simplify calculations and eliminate errors caused by manual formulas;
-
unify inconsistent inputs like dates, units, and formats;
-
generate dashboards that not only visualize data but also explain patterns and suggest next steps.
Challenges
- mapping natural language inputs into reliable database actions;
- making AI-driven outputs transparent enough that bookstore staff could trust them;
- creating dashboards that felt lightweight and approachable, while powerful enough for large catalogs and sales datasets.
Results
- faster catalog onboarding thanks to automated column enrichment;
- consistent data across records, reducing manual clean-up;
- accessible analytics for non-technical staff through natural prompts;
- smoother workflows with AI filling gaps between CRM and CMS systems.
Autofill of data
Managing large catalogs means filling in endless records: categories, authors, publishing houses. Doing this manually is slow and prone to mistakes.
We designed an AI-powered autofill to handle repetitive entry directly in the table. Instead of typing cell by cell, users can select a column and let AI enrich missing values based on existing records or external references:
-
column enrichment prompts: AI suggests missing details like book categories or publishing houses with one click;
-
smart references: users can specify which field (e.g., book name) should act as the reference point;
-
auto-updates: when reference data changes, the system can auto-update linked fields, keeping records consistent over time.
For bookstore managers, this means faster onboarding of new titles and cleaner datasets that can be trusted across the whole system.
Description generation
Writing product descriptions for hundreds of books is time-consuming and inconsistent. Tone varies, details get missed, and external tools are often needed. With conversational AI, users can now generate descriptions inside the system, without relying on external writing tools.
-
inline AI writing: within the description column, users can select “Write for me” or create based on an example;
-
tone & length control: users can adjust messaging style to fit categories (e.g., professional, curious, emotional);
-
error prompts: if a key field (like book title) is missing, the system highlights it before generating.
AI-assisted calculations
Working with complex formulas often excluded non-technical staff and left room for human error.
Our solution: natural language calculations powered by AI. Instead of memorizing formulas, users simply describe what they need in plain words (e.g., “Calculate product performance index”), and the system generates the correct formula instantly:
-
formula preview: the AI displays the generated formula and an explanation of how it works;
-
inline integration: results are inserted into the relevant table column, so users can immediately see the output in context.
Data unification
Inconsistent formats are one of the most common data headaches. Dimensions in inches mixed with centimeters, or weights in pounds next to kilograms, make analysis frustrating.
The AI data management tool detects these inconsistencies automatically and prompts users to unify them into a consistent format:
-
proactive detection: the system surfaces mismatched units with a friendly notification;
-
one-click standardization: users can convert all entries into a chosen system (e.g., metric) without touching formulas;
-
adaptive automation: fields update dynamically when new references are added, keeping datasets clean without extra effort.
AI-driven dashboards
Dashboards are often either too complex for non-technical users or too shallow to provide real value. The challenge was to make visualization both conversational and actionable. This AI-driven dashboard builder turns plain requests into clear visualizations and explanations:
-
natural prompts: users type instructions like “Create a dashboard with revenue, retention rate, and leads by source;”
-
smart layout generation: the system selects the most suitable chart types and assembles widgets automatically;
-
inline analysis: each chart comes with a short AI-generated explanation highlighting trends and correlations;
-
guided exploration: the assistant suggests follow-ups and generates new views.
we reply under 24 hours.
free
session
