Pranav
Modem
Data Artificer. I build the pipelines that move Fortune 500 audience data at inMarket. On my own time I ship AI systems that trade and teach without me.
About me
I build data platforms at inMarket. I own the audience data delivery platform end to end, and before that I spent three years on the ingestion behind LCI, inMarket's closed-loop attribution platform. I care about the unglamorous parts that make pipelines trustworthy: schema evolution, data quality checks, and observability that turns hours of root-cause hunting into minutes.
Relics of battles won
Experience
Every stop, a story. Every role, a relic. — tap a stop to explore
Big Data Solutions Engineer · inMarket
I own the audience data delivery platform end to end: transformation pipelines on Databricks and Delta Lake tuned to tight performance budgets, the orchestration layer that moves data through its full lifecycle, and the reliability work nobody sees until it breaks. Resumable transfers, connection hardening, data-integrity investigations across the full stack. Earlier I built the config-driven ingestion behind LCI, inMarket's closed-loop attribution platform for Fortune 500 advertisers.
Senior Data Engineer · Capital One
Designed automated monitoring, alerting, and remediation frameworks across AWS inside a locked-down VPC. Manual intervention dropped by roughly 97%. Also led partitioning and resource tuning on distributed PySpark pipelines, which cut processing time about 40% on high-volume financial datasets.
Data Engineer · Capital One
Built distributed PySpark and Spark SQL pipelines on Databricks for high-volume financial data, and orchestrated multi-stage workflows with Step Functions and event-driven Lambda. The layered data validation I wrote fed executive reporting and compliance dashboards.
Data Scientist · Advithri Technologies
A short first stop after grad school, doing applied data science before moving into data engineering.
Graduate Student Assistant · University of North Texas
Built an ML model that scored graduate applicants on SOP text, GRE scores, and GPA. The College of Information used it to make admission decisions more consistent.
Projects
Legendary encounters conquered. These are live systems with real users.
Alpha Intelligence
An autonomous trading system that runs two loops. A 60-second price-action loop detects setups on live Alpaca data, then a 10-minute intelligence loop sweeps 27 OSINT feeds and runs a 20-agent debate (bulls, bears, quants, and a risk manager with veto power) before any trade executes. It learns from every outcome.
ELI5Code
Data structures and algorithms explained like you're five, then flipped into precise technical language with one switch. 80 lessons across 8 modules, every one with its own step-by-step visualizer, plus an AI tutor that honors the ELI5/Tech toggle and a 60-day mastery dashboard.
Automated GenAI video pipeline
Takes short-form video from idea to published upload with no manual steps: Claude and Gemini write, text-to-video renders, neural TTS narrates.
Grad admissions ML model
Scored graduate applicants on SOP text, GRE scores, and GPA. UNT's College of Information used it to make admissions more consistent.
Take what you need, adventurer.
Always glad to talk data platforms, AdTech, or agentic AI.