I’m Doniyor, a data analyst working across commerce, operations, and finance.
About
I study Management at Silk Road University and Data Science at School 21. I started learning programming alongside my degree, knowing it would be difficult, and that was part of the appeal. Most of what I know about SQL, Python, and data analysis, I’ve learned by building my own projects.
Work
Four projects in commerce, operations, and finance: what I analyzed, what I found, and where the evidence ends.
Python · forecasting
On orders I kept aside for testing, my model gave an average delivery promise of 17.91 days, compared with Olist’s 20.94 days. It also missed fewer dates: 3.44% of orders were late, compared with 4.50%.
SQL · marketplace operations
The worst-performing group accounted for 5.7% of one-star reviews. The larger “watch” group accounted for 40.5%. I’d start by helping that group improve its shipping, rather than expecting removals to fix the problem.
Power BI · supply chain
Fishing led revenue, but Garden and Crafts had higher margins. Delivery showed another gap: First Class was faster in days, yet missed its promised date much more often than Standard Class.
Excel · financial analysis
Profitability, liquidity, and solvency improved between 2023 and 2024. Inventory grew faster than revenue, so that was the part I would keep watching.
Skills & learning
Practical tools and coursework behind the work, with public examples where available.
SQL · Python · Power BI · Excel · Power Query · DAX
Sampling · confidence intervals · hypothesis tests · A/B tests · regression · ANOVA
Model validation · classification · tree ensembles · clustering · dimensionality reduction · neural networks
Building with a team
I co-founded AdvoAI with To‘rabek. We built the first prototype at the National AI Hackathon and have developed it through many iterations over time. Today, it’s live at advoai.uz and helping people with legal questions.
My part was engineering the backend (retrieval pipeline: document chunking, hybrid search, and reranking) and designing the landing page.
Study & teamwork
Study, collaboration, and building with others shape the work.
01 / School 21
I study data science at School 21, where learning is project-based and peer-led.
02 / National AI Hackathon
Our team developed the first AdvoAI prototype at the hackathon. The work continued into a live MVP.
03 / AI Leaders
A moment from the AI Leaders nomination with fellow students at Silk Road University.
Get in touch
I’m looking for a data analyst role or internship. I’d be glad to discuss my work and hear about your team.