Curious Soul
Builder · Vaibhavi Mutya
Build things to find
solutions to your curiosity.
solutions to your curiosity.
Hi!! I'm Vaibhavi, an engineer who gets excited about the messy, interesting parts of data work. The dataset hiding a pattern nobody noticed. The pipeline that breaks at 2am and teaches you something new. The Friday night idea that turns into something actually useful.
I studied CS, got curious about AI, research, and ended up building pipelines across more industries than I planned (currently pursuing a Master's in AI on the side). Along the way I figured out I learn best by picking up a real problem and building until I understand it.
I studied CS, got curious about AI, research, and ended up building pipelines across more industries than I planned (currently pursuing a Master's in AI on the side). Along the way I figured out I learn best by picking up a real problem and building until I understand it.
Python
dbt
Snowflake
Airflow
Power BI
Azure
NLP / ML
Statistics
My approach to life
My default response to a hard problem is to build something. Every project here started as a problem I personally needed to solve.
Projects
H1B Sponsor Tracker
Parses 3 years of DOL LCA data to identify cap-exempt employers, ranks by H1B volume, and surfaces careers page URLs through search APIs and Playwright scraping.
PythonPlaywrightSerper API
Read the full breakdown →
Job Market Analyzer
Parses PDF resumes, aggregates postings across US/EU/India via APIs, uses HuggingFace NER, spaCy, and zero-shot classifiers to extract skills, seniority, and visa requirements, then ranks markets by fit.
HuggingFacespaCyPython
Read the full breakdown →
Earthquake Dashboard
USGS GeoJSON feed cleaned through a pandas pipeline, surfaced via a Streamlit dashboard with maps, magnitude/depth correlations, and monthly trends. Containerized with Docker.
PythonStreamlitDocker
Read the full breakdown →
E-Commerce Analytics Pipeline
building
End-to-end batch pipeline ingesting e-commerce transactions into Snowflake, modeled with a dbt star schema, orchestrated via Airflow, with a Power BI dashboard tracking revenue trends, customer segments, and return rates.
SnowflakedbtAirflowPower BI
Stock × Industry × Jobs
building
Connects live stock market signals to BLS employment data to forecast which industries will see hiring growth next quarter. Applies time-series models and correlation analysis across sector returns and employment trends.
ARIMA / ProphetBLS APIYahoo FinanceStreamlit
How I think
I love products. I care about the person on the other end, so before building anything, I try to understand why and how it will actually help them. That means mapping the whole workflow before writing a single line of code: inputs, outputs, usability, dependencies, failure points.
As I build each layer, I work through three questions: why does this work, why might it break, and how can I make it more reliable and cost-efficient. When something fails, I don't paste the error into an LLM and accept whatever comes back. I read the error, understand what it actually means, form my own hypothesis about the cause, and then use AI to pressure-test that hypothesis. The difference matters. One builds understanding. The other just clears the blocker.
As I build each layer, I work through three questions: why does this work, why might it break, and how can I make it more reliable and cost-efficient. When something fails, I don't paste the error into an LLM and accept whatever comes back. I read the error, understand what it actually means, form my own hypothesis about the cause, and then use AI to pressure-test that hypothesis. The difference matters. One builds understanding. The other just clears the blocker.
Writing
How I built a tool to solve my own visa problem
A personal story about two layoffs, the broken H1B playbook, and the data pipeline I built to surface cap-exempt sponsors from messy Department of Labor disclosures.
Why interdisciplinary thinking beats depth in the AI era
Why AI raises the value of broad problem framing, systems thinking, and cross-domain insight more than pure technical depth alone.
Can stock market signals predict where hiring will grow? (Coming soon)
A piece on connecting market moves with labor trends, with the same evidence-based thinking I apply to every data project.
Resume
Data Engineer with experience across medical research, talent intelligence, and computational sciences. MS Computer Science · MS AI (in progress).
Open to
Full-time roles
Contract roles
Open source
Coffee chats
Collaborations
Get in touch
Vaibhavi Mutya · Bay Area · vaibhavimutya.github.io
Built by me, obviously.