Making Sense of Complex Data

From PDFs to Dashboards: Wrestling Chaos Into Clarity

Making Sense of Complex Data
Photo by Ardian Pranomo / Unsplash

The Real Work of Data

Clean, structured data is rare in the wild. Most often, it's messy, outdated, or locked away in hard-to-read formats. That’s where I find my flow—transforming complexity into clarity.

When Energy Data Lives in PDFs

Take the POSOCO project. Every day, India’s power usage data gets uploaded as PDFs—great for archiving, terrible for analysis. So I built a scraper using Python, BeautifulSoup, and Tesseract that parsed these into daily datasets. With just a few scripts, energy trends across India became graphable, actionable insights.

Electricity demand data collected from POSOCO (Python) and charted using ggplot2 (R)

Data Tools for People, Not Just Analysts

At Catalyst Management Services, I worked with field teams collecting WaSH data for NGOs. Instead of fancy dashboards, they needed Excel—and so I designed a full-stack solution in VBA that made their data usable and visual without ever needing to open a database. That mindset—meet the user where they are—is what defines useful data work.

Not Just Data. Stories.

Whether parsing a resume stack with OCR or crunching numbers for MSME energy usage, my job is to uncover the narrative behind the numbers. Tools like pandas, ggplot2, and Power BI are just instruments—what matters is the story we tell with them.

Charting the publically avaialable data on Mario Karts favourite tracks