Software engineer from Nepal. I build backend services and data platform infrastructure, and contribute to Apache DataFusion and Arrow projects. Currently an SDE I at HighRadius on the G4 Data Platform team.

📫 unikdahal03@gmail.com · LinkedIn · Website

• Query connectivity: replaced JDBC/Thrift with Arrow Flight SQL and ADBC, cutting fixed query overhead from ~1.5 s to ~90 ms
• Analytics migration: supported moving analytical workloads from Snowflake to Spark, Iceberg and Kyuubi (~90% lower compute cost on migrated workloads)
• Distributed workflows: built a Saga-based orchestrator across seven microservices; environment recreation went from ~6 months to ~1 month
• Multi-tenancy: fair query admission using Redis-backed leases and tenant-aware routing
• Production debugging: database lock contention, catalog and metadata issues

Merged

• DataFusion Comet: native Iceberg MERGE execution, plus WriteDelta planning and delete-metadata correctness
• Apache Arrow ADBC: Java Flight SQL flightsql:// URI support
• Apache DataFusion: fix for a panic on zero-field structs

In progress

• Native merge-on-read writes in DataFusion Comet
• Row-group-level runtime pruning for Iceberg (iceberg-rust / DataFusion Comet), up to 42x faster on one benchmark query

• redis-java: a Redis-compatible server in Java 25 and Netty. It covers RESP, pipelining, transactions, streams, replication, and 599+ tests.
• Sutine: a full-stack e-commerce platform for the clothing label I co-own.

Java SQL TypeScript Python C++ Rust · Spark Iceberg Kyuubi Arrow Flight SQL Kafka Redis PostgreSQL MySQL · Docker Kubernetes AWS

B.Tech in Computer Science & Engineering (Minor: Financial Economics), KIIT University · CGPA 9.12

🏆 HighRadius Fresher of the Year (HighFlyer 2025)