Devank Uppal

Aug 2026 – present · Intelligent Solutions Group · Ames, IA

data lake my PySpark and SQL scripts pull from, joining telemetry and engineering sources

PB-scaledata lake my PySpark and SQL scripts pull from, joining telemetry and engineering sources

use my dashboards for new and existing products and features

Teams + engineersuse my dashboards for new and existing products and features

• Write PySpark and SQL on Databricks that join telemetry and engineering sources from a petabyte-scale data lake into analysis-ready views, for agricultural R&D and existing products.
• Build Tableau dashboards and reusable data views that product teams and individual engineers use for reliability analysis and feature decisions.
• Support Right to Be Forgotten data compliance.

Apr 2026 – present · St. Louis, MO

fiakominds.ai, built and launched, indexed on Google

Livefiakominds.ai, built and launched, indexed on Google

new users get their login details the moment their account is created; someone used to email them by hand

Automaticnew users get their login details the moment their account is created; someone used to email them by hand

• Built and launched the company website, fiakominds.ai: requirements, code review, deployment, SEO and Google Search Console.
• Built an internal React platform on Supabase with Microsoft Entra SSO that creates accounts, assigns licenses and sends the new user their login details automatically, a step someone used to do by hand.
• Now building the data ingestion for PulseIQ, the company's customer sentiment tool.

May – Jul 2026 · 28-person migration team · St. Louis, MO

In production

Salesforce extraction per run, was 45 min to several hours (~2 hrs of manual work)

17 minSalesforce extraction per run, was 45 min to several hours (~2 hrs of manual work)

match with the manual export, across 464K+ records per run

100%match with the manual export, across 464K+ records per run

• Automated the Salesforce extraction that gated every migration run (AWS Glue → Salesforce API → S3): ~17 min per run instead of waiting 45 min to several hours on one person, matching the manual export 100% across 464K+ records (154K accounts, 310K opportunities). In production.
• Coordinated project management for the 28-person migration team: leadership status reports, cascading timelines, and deployment handoffs between the business, the vendor and DevOps: how enterprise software gets shipped, beyond the code.
• Also built and tested; not in production when I left: a parallel contract-database extraction (credentials in Secrets Manager via Terraform) and a Bulk API flag update with an approval step.
• Built a pipeline monitoring web app to replace DevOps' command-line tool: anyone on the team could start a run, watch each step's live status and approve or reject by email, with an audit trail (built and tested; not in production when I left). Deployed it with DevOps on EC2 inside the VPC, working through security groups, inbound/outbound rules and IAM roles.
• Designed the security and access model for an MCP-based AI architecture (master orchestration agent, document-level permissions); approved by the business.