Omkar A.

Most data problems don't start with bad analysis. They start with bad data. * A website changes its structure. * An API returns inconsistent records. * A PDF contains information buried across hundreds of pages. * Five different sources use five different formats. * A spreadsheet is updated manually every morning. * Eventually, someone gets stuck cleaning everything before the data can be used. That's the part I solve. I'm a Python Data Engineer focused on building the systems behind reliable data: from collection and extraction to processing, validation, storage, analytics, automation, and creating AI agents. I've worked on data problems involving financial markets, SEC filings, real estate, healthcare products, public web data, APIs, Excel, and large historical datasets. Some examples of what I've built: # SEC Filing Intelligence: extracting and processing financial filings into structured research data. # Trading Analytics Pipelines: collecting market data, transforming it, and turning it into usable analytics. # Real Estate Data Systems: collecting property data from multiple sources and creating structured market datasets. # Automated Research Workflows: replacing repetitive collection, comparison, analysis, and reporting with automated pipelines. # Web Scraping Systems: extracting difficult or changing website data and converting it into clean datasets. # AI Data Workflows: using LLMs where they actually add value: classification, extraction, enrichment, analysis, and workflow decisions. * I don't approach these projects as "just scraping" or "just ETL." The real question is: What needs to happen to the data after it is collected? If the answer involves cleaning, deduplication, validation, historical storage, scheduled updates, APIs, alerts, analytics, or downstream automation, that's where I can build the complete workflow. My core stack Python · Pandas · SQL · APIs · Web Scraping · ETL/ELT · FastAPI · Databases · Excel · Financial Data · SEC/EDGAR · AI/LLM APIs · Automation I care about the things that make a data system usable in the real world: accuracy → reliability → maintainability → automation. If you already know exactly what you need, send me the requirements. If you only know "we have this data problem, and we're doing too much of it manually," that's fine too. Tell me what you're trying to accomplish. I'll help map out the data flow and the simplest way to build it.