Triaging Customer Support Tickets
Gemini labels 44 grocery delivery support tickets in fixed terms, and the team's own rules route them. All 44 were routed, and urgency matched the team on 10 of 12.
Anjali
I build data pipelines, and the generative AI workflows that run on them. 5 years in data analytics, now in Microsoft Fabric and Databricks, building LLM pipelines that check every AI answer before anyone relies on it.
- Microsoft Fabric
- Databricks
- Generative AI
- PySpark
- SQL
Open to Data Engineering and Generative AI roles · Available to join immediately · Get in touch
- 5
- Years in data analytics
- 80%
- Less manual file processing
- 55%
- Less source query load
Case studies
Each one documents the decision, not just the result, including the scenarios that tested whether the design would hold. All are practice projects.
Unified Commerce Analytics
Replaced a database that needed five joins for one revenue number with a Fabric Lakehouse and a Direct Lake star schema, without touching the point-of-sale system once.
Workforce Intelligence: Azure SQL to Lakehouse
Replaced HR's manual SQL exports and Excel cleaning, which delayed reports by about 48 hours, with a Fabric pipeline that quarantines duplicates, invalid salaries and missing hire dates.