background Layer 1

Empowered Employee: Transforming Work with Generative AI

Customer
ABB Bank
Project manager on the customer side
Nilufar Shafiyeva
Senior Product Owner
Year of project completion
2025
Project timeline
March, 2025 - December, 2025
Project scope
3675 man-hours
Goals

The project aims to empower ABB Bank employees with a Generative AI-driven assistant that simplifies daily operations, improves access to internal information, and automates repetitive requests. The solution enables staff to retrieve data, generate applications, and receive accurate responses in seconds, regardless of time or location. Its goal is to build a seamless digital workspace where employees can focus their energy on strategic work while AI handles administrative routines, fostering a data-driven and innovative culture within the bank.


Project Results

Since launch, 3 300 employees (87.2 % of ABB’s staff) have adopted the assistant, submitting over 55 000 queries with an accuracy rate of 87.7 %. The solution supports 12 domains, with HR accounting for 52 % of interactions. Around 1 600 employees use it monthly, generating 5 000–10 000 requests. Email traffic to HR has dropped by 70 %, and response times that once took minutes now take seconds — resulting in higher productivity, faster internal service delivery, and a more efficient workplace.


The uniqueness of the project

This is the first enterprise-level Generative AI assistant developed exclusively for ABB Bank employees. Unlike traditional chatbots, it integrates 12 internal domains — HR, Operations, Compliance, Agile, IT Infrastructure, Procurement, FISDD, MMKD, NHD, MUD, and AI-Khan product knowledge — into one secure platform. The solution combines RAG pipelines, ERP and ManageEngine APIs, and SQL + Visualization Assistants within a single Web App interface. Its ability to interpret intent based on departmental context and deliver actionable results makes it a pioneering multi-domain GenAI solution in the regional banking sector.


Used software

The system integrates OpenAI API for language generation with ABB’s enterprise infrastructure. ERP and HR systems provide real-time data through API links; ETL pipelines feed structured data from PostgreSQL databases; a RAG architecture retrieves contextual knowledge. ManageEngine integration enables process automation, while Qlik dashboards deliver analytics. SQL and Visualization Assistants are embedded in the same interface for data queries. Gleap supports monitoring, and the Web App runs securely on ABB’s internal network, ensuring full compliance with banking data-protection standards.


Difficulty of implementation

The implementation was complex due to multi-system integration and advanced AI logic. Building APIs with ERP, HR, and ManageEngine platforms, setting up ETL data pipelines, and implementing a RAG model for contextual accuracy required deep coordination between AI and infrastructure teams. Intent detection and routing were customized for 12 internal departments, each with unique data tables and workflows. Ensuring data security, maintaining compliance with banking regulations, and driving employee adoption simultaneously made the project both technically and organizationally challenging.


Project Description

The project introduces an enterprise-level Generative AI solution designed to support ABB Bank employees in their daily operations. Before its implementation, internal processes such as checking vacation balances, submitting remote-work or replacement-day requests, and finding HR or compliance information were performed manually via email or through separate internal systems. Searching for the correct forms or data across multiple platforms was time-consuming and often required HR or IT support.

The new solution replaces these fragmented interactions with a single, intelligent virtual assistant powered by Generative AI. Employees can now simply type their questions in natural language — for example, “What is my remaining vacation balance?” or “Submit a distant-work request for tomorrow” — and receive instant, accurate, and context-aware responses. The assistant not only provides answers but also executes requests automatically through integrations with internal systems.

Technically, the platform combines OpenAI’s large language model API with ABB’s enterprise ecosystem. It integrates with ERP systems for employee and operational data, applies ETL processes to gather structured data from PostgreSQL databases, and uses a RAG (Retrieval-Augmented Generation) pipeline to retrieve contextual information for precise answers. ManageEngine and HR platforms are connected for automated form execution and approval routing. A complex intent-detection and multi-domain routing mechanism enables the assistant to correctly interpret employee needs across HR, Compliance, Operations, IT, and Procurement domains.

The system also incorporates Gleap analytics to track interactions and performance metrics, while Qlik dashboards provide visualization of usage data and KPIs for management insight. All components operate within ABB’s secure on-premise environment, fully compliant with banking data-protection requirements.

As a result, what previously required multiple emails or manual steps can now be accomplished in seconds. The assistant significantly reduces administrative workload, improves employee experience, and accelerates internal service response times. It embodies ABB’s strategic vision of creating an AI-empowered, data-driven workplace where technology simplifies processes and enables employees to focus on higher-value tasks. The project is a key step in ABB Bank’s broader AI transformation journey, establishing a scalable foundation for future Generative AI innovations.

Project geography

The project has been deployed across ABB Bank’s entire internal ecosystem in the Republic of Azerbaijan, covering all head-office units and regional branches. It serves employees in HR, Operations, IT, Compliance, Procurement, and other domains nationwide. The assistant is accessible via ABB’s secure intranet as a Web App, ensuring unified performance across all locations. Designed for scalability, it will later extend to other ABB entities and potentially to regional subsidiaries, supporting the vision of a unified, AI-empowered organization.


Additional presentations:
Employee AI.pdf
We use cookies for analytical purposes and to deliver you the best experience with our website. Continuing to the site, you agree to the Cookie Policy.