Data Science & Intelligence Projects
I build systems that turn large, messy datasets into useful signals for understanding markets, competitors and customer-facing products.
My projects sit at the intersection of software engineering, data engineering, analytics and data science. I enjoy working on problems where collecting the data is only the beginning: the real challenge is deciding what information matters and turning it into something useful.
Competitor Intelligence Engine
Turning marketplace data into competitive intelligence.
Problem: raw product listings don't tell you who your competitors are or how you compare. Demonstrates: data engineering, data modeling, analytics thinking, business problem solving.
Read the Case StudyPrice Intelligence
Turning marketplace prices into actionable pricing information.
Problem: identifying price violations reliably, not just collecting prices. Demonstrates: root-cause analysis, data pipeline design, requirements analysis, business-to-engineering translation.
Read the Case StudySearch Intelligence
Understanding brand visibility and competitive movement across search results.
Problem: raw keyword rankings are difficult to interpret on their own. Demonstrates: analytics, data visualization, competitive intelligence, decision-oriented data presentation.
Read the Case StudyAutomated Product Matching Pipeline
Reconciling listings across competing retailers.
Problem: the same product has a different title, photo and SKU on every retailer's site. Demonstrates: entity resolution, precision/recall tradeoffs, data validation at scale.
Read the Case StudySARA
A reusable data retrieval architecture: the engineering foundation behind the intelligence work above.
Overview: a modular architecture for retrieving and processing data from 22 sites, separating responsibilities instead of one monolithic crawler. Demonstrates: system architecture, modular design, reusable engineering.
Read the Case StudyModular Web Scraping Pipeline
Production-scale, multi-site extraction with YAML-driven configuration.
Problem: every new site meant a new one-off script. Demonstrates: pipeline design, configuration-over-code thinking, the engineering foundation the rest of this work sits on.
Read the Case StudyData Quality & Field Prioritization
Finding which data actually matters.
Problem: a recurring QA process was flagging issues on fields with no real business value. Approach: categorized fields into P0/P1/P2/On-demand/Not-useful, shifting focus from "fix every field" to "fix what's worth fixing."
Read the Full StoryOn Data Pipeline Architecture
A significant part of my engineering work has involved building the layers between raw external data and the information consumed by downstream systems: collection, fetching, extraction, validation, transformation, storage, analytics, business intelligence.
That work taught me that data engineering isn't only about moving data from one place to another. The difficult part is deciding what should be collected, what should be ignored, how data should be structured, how quality should be validated, and how downstream consumers will actually use it. Every project above is a different answer to some part of that question.
Working on a similar problem?
Happy to talk through pipeline architecture, data quality, or turning messy marketplace data into something useful.