About Me

Data Science | Artificial Intelligence | Business Intelligence

I'm a systems thinker who thrives at the intersection of data, intelligence, and impact. With a background spanning data science, business intelligence, and applied AI, I approach every challenge with structured problem-solving and a focus on scalable, real-world solutions. Whether I’m designing an analytics framework or building multi-agent AI systems, I aim to create tools that are intelligent, reliable, and aligned with business needs.

I believe in servant leadership, personal accountability, and fostering cultures built on trust. I aim to be the kind of teammate who empowers others, takes responsibility before placing blame, and leads by example—not by title. My work is grounded in first-principles thinking, the scientific method, and agile iteration. I value clarity, modularity, and experimentation—qualities that drive meaningful innovation without sacrificing long-term stability. I’m always learning, always refining, and always building systems that turn ideas into action.

My Natural Languages: English, Serbian, Bosnian, Croatian, Bulgarian, Rusian.

My Machine Languages: Python, C#, R, Julia, JavaScript, Java.

View Portfolio View Resume

What I do

I design and deliver full-stack data and AI systems that span engineering, intelligence, and automation. My data engineering work focuses on building reliable, scalable ETL/ELT pipelines using Python, PySpark, SQL, and orchestration platforms like Fivetran, Stitch, Funnel, and DBT, deployed across cloud-native stacks including Databricks, Snowflake, Azure Data Factory, and Microsoft Fabric.

My AI work focuses on developing intelligent agents and predictive solutions using tools including Mosaic AI and Azure ML, along with frameworks such as LangChain, LlamaIndex, and PydanticAI. I design systems around principles like A2A (agent-to-agent communication) and MCP (modular, composable pipelines), enabling scalable, testable architectures for both enterprise and experimental use.

My focus is always on end-to-end solutions: architecting systems where raw data becomes intelligence, and intelligence becomes action, whether through dashboards, models, or autonomous agents built for real-world impact.

Python Programming

With a deep understanding of Python's versatile capabilities, I engineer sophisticated data structures and algorithms. My proficiency in Python is demonstrated through automation, data analysis, and building scalable ETL pipelines, contributing to robust data solutions.

Data Science

My data science expertise lies in extracting meaningful insights from complex data sets. Using statistical analysis and predictive modeling, I uncover trends and patterns that inform business strategies and drive innovation.

Artificial Intelligence

Harnessing modern AI frameworks and platforms—including LangChain, Pydantic-AI, LlamaIndex, Mosaic AI, and Azure OpenAI, I develop interactive systems powered by LLMs and vectorized search. These technologies enable designing advanced AI agents that mimic human reasoning, streamline workflows, and unlock new product capabilities.

Data Engineering

Using Azure services in combination with Databricks and Snowflake, I architect and maintain scalable, reliable data pipelines based on a clean medallion architecture. By leveraging PySpark and Spark SQL, I ensure high data quality, consistency, and accessibility, supporting seamless data flow from extraction to preparing data that’s ready for AI development.

MSSQL | MySQL | SQLite | Postgresql

My SQL expertise spans across MSSQL, MySQL, and SQLite. I design, query, and manage databases, ensuring data integrity and performance. My work supports complex data ecosystems and business intelligence applications.

Power BI | Tableau | Sisense | Looker

Proficient in Power BI, Tableau and Sisense, I transform data into rich visuals and interactive reports, enabling stakeholders to grasp complex scenarios and make data-driven decisions. My dashboards reflect a synthesis of clarity, accuracy, and actionable insights.

Google Analytics & Big Query

Utilizing Google Analytics and Big Query, I harness the strength of web analytics and large-scale data processing. This allows me to make data-informed decisions to enhance user engagement and business performance

Robotics

Combining Python with Raspberry Pi, I delve into the innovative world of robotics. I develop smart devices and systems that can sense, process, and act upon data, opening avenues for intelligent automation and real-world problem-solving.

Work Experience

Analytic Vizion

As a Solution Architect and Data Scientist at Analytic Vizion, I work closely with existing clients to deliver tailored data science solutions while also contributing to the growth team by supporting business development and onboarding new customers. I apply my expertise in data science, data engineering, and AI development to solve complex, domain-specific challenges and help organizations unlock the full potential of their data on modern infrastructure.

analyticvizion.com

CPAP.com

As a Data Scientist and Business Intelligence Architect, I lead comprehensive analytical studies, working with structured and unstructured data, using mathematical and statistical methods to interpret and analyze business and customer-related data. I performed Customer Lifetime Value Analysis, forecasted item sales, and developed Power BI-based reports for all company teams. I am responsible for conducting experiments to discover new solutions to data challenges and for building data environments according to stakeholder requirements.

cpap.com

Ash Projects

As a Data Scientist, I developed ETL processes for diverse customer projects using Python, Django and Flask frameworks. I built real-time dashboards using Sisense and Power BI and crafted a customer behavior system using time series analysis for occupancy and transaction predictions. My great effort was placed on creating web applications with features such as facial recognition, mask detection, gender detection, age detection, and distance tracking, by applying machine learning for computer vision on platforms like Raspberry Pi and Jetson Nano equipped with Intel RealSense cameras.

ashprojects.com

HDSupply

As a Data Scientist, I employed machine learning techniques for data mining and modeling. I crafted a series of web scrapers and API scripts for data enrichment utilizing custom-developed similarity algorithms. I monitored performance processes using Python, SQL, and R, and served as a primary data source for scorecards and KPIs. Additionally, I led development projects, including company acquisitions, taxonomy, and time series analysis for future forecasting.

hdsupply.com

FLIR/BrickStream

As a Solution Analyst, my role involved analyzing customer data, leading project management efforts, designing site layouts, configuring and validating various video sensors ensuring accurate data delivery. Furthermore, I onboarded and trained new employees and managed a team overseas and played a key role in deploying real-time dashboards for major retailers. Notably, I led a project for ASDA that spanned 320 stores, coordinated international teams, and provided technical guidance for specialty retail projects.

flir.com

The Home Depot

As a Business Data Analyst, I specialized in data analysis and problem-solving, utilizing Excel, and SQL Server to draw insights and solutions. My responsibilities extended to training, contract negotiation, and supporting a wide range of programs and teams, managing the MicroStrategy and SharePoint, conducting labor studies, developing KPIs, integrating automation, and improving efficiency.

homedepot.com

View Full Resume

Featured Projects

Hierarchical Autonomous Agents

The Hierarchical Autonomous Agent Swarm (HAAS) is an advanced system composed of three interconnected agents, each specialized to perform tasks related to querying SQL database and analytics. This system enables efficient handling of natural language prompts to execute and analyze data from SQL database.

GitHub

View Case Study

RAG - AI Agent

Application designed based on retrieval-augmented generation model, using Langchain framework and Streamlit interface to provides question-answering capabilities using Large Language Models (LLM). It allows users to upload documents, process them, and use a LLM to answer questions based on the content of these documents.

GitHub

View Case Study

Mars Rover

An innovative robotic platform featuring multi-terrain mobility, equipped with sensory and control systems, designed for versatile applications ranging from environmental monitoring to automated material handling.

GitHub

View Case Study

Synthetica app

GUI-based application designed to interact with OpenAI's GPT and DALL-E models, providing users with the ability to generate text and image responses to prompts. It features a modern interface, history management of prompts, and secure credential handling for API interactions.

GitHub

View Case Study

View Portfolio