The question determines the method
I establish what needs to be explained, estimated or decided, and which evidence can support an answer. Metrics and methods follow the specific problem.
AIQU Systems brings applied research, advanced analytics and analytical development together in a practice led by Cristina Sârb.

Founder, AIQU Systems · PhD in Mathematics
I work directly on AIQU projects, from understanding the question and investigating the data to method development, validation and delivery.
Results take a usable form: evidence-based findings, a model, a repeatable workflow or a working internal tool.
My experience connects mathematical research with more than a decade of applied analytical work across distribution, software and real estate — forecasts, ranking methods, statistical analyses, data workflows and internal tools.
I focus on problems where the intended outcome is clear but the approach still needs to be developed: what to measure, which data to use, how to model the problem and how to assess whether the method works.
I completed my mathematics degree and master’s in Applied Mathematics at Babeș-Bolyai University, where I earned my PhD in Mathematics in 2009.
Research in approximation theory and statistical convergence shaped an approach I still use: define the problem precisely, make assumptions explicit and establish when a result holds.
I analysed sales, inventory, product assortments and customer purchase histories to support replenishment and commercial decisions.
I also developed a Windows application that proposed orders using sales history, field inputs, seasonality and commercial rules. The commercial director could review and adjust each proposal before submitting the order.
The tool remained in use after I left.
At Yardi, I owned analytical projects from question formulation through to conclusions and recommendations. The work covered forecasting, property comparables, ranking, conversion, cohorts, statistical evaluation and digital product research.
I integrated data across SQL Server, Oracle, APIs and large files, including datasets with tens of millions of records. I developed SQL processes, extraction tools and automated reporting workflows that made analytical work repeatable.
Over time, responsibilities expanded across several products and business divisions, including analytical and technical support for other analysts.
As a co-founder of AIQU Media, I work on methodology, analysis and technical workflows for original research intended for publication.
Through AIQU Systems, I bring the same combination of research, analytics and implementation to companies and organisations — from focused investigations to methods that need to become part of recurring operations.
Doctoral research on approximation processes — convergence conditions and the character of approximation error — was published as Cristina Radu between 2007 and 2012.
That training informs how I examine assumptions, test method behaviour and delimit the conclusions the results support.
The steps are not strictly linear. Testing may send us back to the method or the data.
I establish what needs to be explained, estimated or decided, and which evidence can support an answer. Metrics and methods follow the specific problem.
I check the data, intermediate results and method behaviour. Depending on the project, that means historical comparisons, testing on separate periods or investigating cases where the output differs from expectations.
An analysis may support a decision. A method may become a recurring calculation, an automated workflow or an internal tool. The deliverable takes the form your team needs to use it.
You speak with the person investigating the data and developing the solution. Method choices, findings and adjustments are discussed throughout the project.
I choose data and development technologies to match the problem, the data and the intended use.
AI can assist with coding, extraction and exploring alternatives. Method selection, interpretation and final validation remain my responsibility.
Tell me about the question, the available data and how you need to use the result. We can establish what needs investigating and the appropriate form of delivery.