Wall-Street.ro invited dr. Marian Călborean, Managing Partner of OPTI Software, to contribute a timely perspective on the IT&C sector in the Romanian economy. The opinion piece published on May 13, 2026 addresses the paradox of digitalization in the Romanian economy and AI integration: the economy runs on legacy systems while consumers and commercial partners expect e-commerce speed.
"The fundamental problem is not the ERP software itself. These systems were designed to be the transactional backbone of the company: rigid, precise, immutable. Asking an ERP to function as a modern recommendation engine or a sales assistant is like asking an accountant to write advertising slogans while doing the balance sheet."
dr. Marian Călborean, Managing Partner OPTI Software
Key ideas of the editorial:
- The importance of existing data: Companies that dominate the economy have in their systems products (tens, hundreds of thousands), customer history (tens, hundreds of thousands) and their own infrastructure. Integration must start from connecting this data with intelligent systems.
- Sales and operations teams spend a lot of time searching ERP for data that should be immediately available. Sales agents lose between 30 and 60 minutes to generate a complex offer.
- Using hybrid architecture for integrating AI and AI agents: The solution is not "replacing the ERP", a traumatic, extremely costly and risky process that few Romanian companies can afford
The new normal and the "open architecture"
Marian emphasized that commercial flows that rely today on operators across different channels (e.g.: WhatsApp) can be modernized:
"A multimodal system (Vision AI) "reads" the photo, identifies products regardless of handwriting or regional naming, checks stock in real time in ERP and generates the draft offer in 3 seconds"
To demystify the necessary architecture, OPTI recently launched the series of AI Technical Guides for B2B 2026, which details "at the grassroots level" AI integration in distribution and production flows.
The original research comprises 6 open-source guides, dozens of press echoes and was awarded by ANIS in May 2026 with the distinction of Best R&D Program
