REFERENCES

In every unique situation, we strive to find the solution that best fits the specific needs. The projects presented below illustrate cases where we achieved the set goals in close cooperation with our clients. We are truly grateful for every collaboration and the trust we receive. However, an even greater honor is seeing our clients' satisfaction, which serves as both confirmation and motivation for our team.

01

Company:

MBH Bank Nyrt.

Sector:

Finance

Challenge

The 2023 merger of MKB Bank, Takarékbank, and Budapest Bank created a “super-bank” with approximately two million customers and nearly 5,000 employees. To preserve its premier service quality, MBH needed to boost operational efficiency—generative AI offered a compelling path.

Solution

Joint discovery pinpointed high-impact use cases in HR, communications, marketing, and legal. An internal AI chatbot („AIDA”) now answers staff questions, drawing on policies and knowledge bases and citing the exact paragraph for instant validation.

Result

Employees save significant time—needed information is always just a chat away, eliminating manual searches.

02

Company:

META-INF Ltd.

Sector:

IT

Challenge

The Atlassian specialist wanted to harness emerging Generative AI tools in daily work and product development.

Solution

Optimaze delivered a two-part, tailor-made workshop: Part 1 explored LLM capabilities through case studies; Part 2 mapped concrete AI use-cases in META-INF’s software and consulting services. 

Result

Participants from multiple departments could immediately envision AI integration; several ideas are already in design phase. 

03

Company:

State Asset Registry Ltd.

Sector:

Public Sector

Challenge

After migrating to new databases, mismatches between old and new tables were suspected—manual comparison would be slow and error-prone. 

Solution

An RPA bot now compares legacy and new tables in minutes, flags structural differences and extracts data errors. 

Result

Staff know exactly which rows need correction; task duration dropped from days to hours and error rates were minimised.