AI-driven Acceleration of UX Research at TBC Bank Uzbekistan
- Customer
- TBC Uzbekistan
- Project manager on the customer side
-
- IT Provider
- Kotib Research
- Year of project completion
- 2025
- Project timeline
- July, 2025 - October, 2025
- Project scope
- 15 automated workstations
- Goals
-
The project aimed to increase the speed and consistency of user research by automating transcription and analysis of customer development interviews.
Previously, TBC’s UX team manually transcribed and coded hours of interview recordings, which slowed insight delivery.
Kotib Research automated this process, providing structured transcripts, semantic grouping, and AI summaries of user feedback. This allowed researchers to focus on hypothesis validation and decision-making instead of routine documentation.
- Project Results
-
Results:
– Average time to process one interview dropped from 3 hours to 20 minutes.
– Research backlog reduced by 70%.
– Analysts gained access to AI summaries and keyword clusters across dozens of sessions.
– Improved speed of iteration in product design and customer journey optimization.This project showed how AI can transform qualitative research into a scalable, measurable process — a major leap for UX maturity in the Uzbek banking sector.
The uniqueness of the project
The project is one of the first examples in Central Asia of applying AI to accelerate qualitative UX research in the banking industry.
The Kotib Research system was customized for Uzbek, Russian, and English speech — a typical multilingual setting in TBC’s interviews.
The solution not only delivered accurate transcriptions but also extracted key insights, pain points, and recurring themes across multiple sessions. This helped bridge the gap between user research and product decision-making, setting a new standard for evidence-based innovation in the region’s financial sector.
- Used software
- Kotib Research (AI transcription and analytics platform), Google Cloud Speech-to-Text, OpenAI GPT models for summarization
- Difficulty of implementation
-
The key challenges included adapting transcription for mixed-language data. The AI model was trained with local accents and financial terminology to ensure accurate output.Continuous collaboration between Kotib’s engineers and TBC’s UX department ensured smooth workflow integration.
- Project Description
-
TBC Bank Uzbekistan, one of the country’s most innovative financial institutions, conducts continuous user research to enhance its experience. However, the process of transcribing and analyzing dozens of customer interviews each month was time-consuming and prone to inconsistency.
Implementation:
– Integrated Kotib Research into the UX team’s research workflow.
– Connected the platform to internal storage for automatic upload of interview recordings.
– Implemented transcription models adapted to Uzbek-Russian mixed speech.
– Configured AI modules for tagging key topics, customer emotions, and usability issues.
– Provided dashboards for tracking recurring feedback themes. - Project geography
- Uzbekistan (Tashkent)