Sapana Pokharal

Final year B.Tech CSE student, building data pipelines on Databricks & Azure
SIH'24 WINNER

I work at the raw → refined → curated layer of data — turning messy sources into clean, reliable, analytics-ready datasets with Medallion Architecture on Databricks. Most of my time lives inside Spark, Delta Lake, and Unity Catalog, with a growing focus on layering Mosaic AI (RAG, embeddings, LLM integration) on top of solid data foundations.

Currently an Azure Integration Intern at Protiviti, working hands-on with Azure Databricks in a live enterprise environment — bridging coursework and real infrastructure.

Data Engineering DatabricksPySparkSpark SQL Delta LakeMedallion ArchitectureUnity Catalog ETL PipelinesAzure Databricks
AI & Data Platform Mosaic AIFoundation Model APIRAG EmbeddingsAI GatewayMLflowLLM Integration
Cloud & DevOps AWS EC2LambdaS3VPC SNSSESSQS
Networking & Security TCP/IPOSI ModelSubnetting SSH / TELNETDNS / DHCPFirewallsVPN Concepts
Languages PythonSQL
Tools & Platforms Linux (Ubuntu)Cisco Packet TracerEVE-NG
Azure Integration Intern — Protiviti, India
Jun 2026 — Present
  • Worked on Azure Databricks-based data engineering solutions involving ingestion, transformation, and processing using PySpark and SQL.
  • Developed and optimized ETL pipelines following Medallion Architecture (Bronze, Silver, Gold) with Delta Lake.
  • Designed SQL queries and analytical datasets to support reporting and business insights.
  • Built AI-powered data agents using Databricks AI capabilities for natural language querying.
  • Worked with Mosaic AI, LLMs, and RAG workflows for GenAI-based data solutions.
ProjectStackLink
1
Mosaic AI Sales Data Project
End-to-end sales pipeline on Medallion Architecture with logging, embedding generation, and AI processing via Databricks Mosaic AI.
DatabricksMosaic AIEmbeddings
2
Sales Medallion Pipeline Optimization
Performance-tuned sales pipeline with broadcast joins, fact/dimension modeling, Autoloader, and incremental processing.
PySparkDelta LakeAutoloader
3
Mosaic AI
Databricks notebooks demonstrating GenAI workflows — Foundation Models, Embeddings, Vector Search, RAG, AI Gateway.
RAGVector Search
4
Data Ingestion using Lakeflow Declarative Pipelines
Automated ingestion pipeline using Databricks Lakeflow Declarative Pipelines with Medallion Architecture.
LakeflowAutomation
5
HR Hierarchy App
Role-based HR hierarchy application built on Databricks.
DatabricksPython
6
Lakeview Dashboard
Interactive sales analytics dashboard on Delta Lake and Medallion Architecture using Databricks Lakeview.
Delta LakeDashboards