Palash Sharma

Software Engineer

Sunnyvale, California, United States9 yrs 7 mos experience

Highly Stable

Key Highlights

Expert in database server technologies and query optimization.
Holds a US patent in SQL Plan Management.
Proven track record in enhancing Oracle's In-Memory Technologies.

Stackforce AI infers this person is a Database Technologies expert with a focus on High Performance Computing.

Contact

Skills

Core Skills

Query OptimizationHigh Performance Computing (hpc)Sql Plan ManagementShardingSql QuarantineMpi

Other Skills

AlgorithmsAutomationBenchmark validationCC and C++C++Computer ArchitectureDatabase OptimizationDatabase PerformanceDatabase ScalabilityDatabasesFortranJavaJavaScriptLinux

About

Senior Member of Technical Staff at Oracle America Inc. with deep expertise in database server technologies, query optimization, in-memory systems, and HPC. Over nearly a decade at Oracle, drove advancements in In-Memory Technologies, SQL Plan Management (earning a US patent), Sharding, and SQL Quarantine. Looking to leverage deep technical expertise and leadership skills to drive innovation.

Experience

Oracle

4 roles

Senior Member of Technical Staff

Nov 2021 – Nov 2025 · 4 yrs · On-site

o HPK4SQL and VectorDB: Engineered performance enhancements for Oracle In-Memory, Oracle Exadata queries by designing and implementing efficient algorithms that leverage SIMD vectorization (AVX2/AVX512) and parallel execution, achieving significant speedups validated against SSB and TPC-H benchmarks.
o Developed and tested high performance kernels for efficient parsing and loading of Sparse Vector embeddings into Oracle Vector database.

Query OptimizationHigh Performance Computing (HPC)AlgorithmsC and C++SQL

Senior Member Of Technical Staff

Promoted

Jan 2019 – Oct 2021 · 2 yrs 9 mos · On-site

o SQL Plan Management: Spearheaded SQL Plan Management automation, ensuring consistent database performance by enforcing verified execution plans.
Patent: US20240054127A1, Automatic prevention of plan regressions.
o Oracle Sharding: Implemented multi-row DML support for Oracle Sharding, enhancing database scalability.
o Strengthened query optimization ecosystem by maintaining and extending modules including RDBMS Optimizer, SQL Test Case Builder, SQL Diagnosis and Repair, SQL Quarantine, and SQL Plan Management.

SQL Plan ManagementShardingQuery OptimizationPL/SQL

Member Of Technical Staff

Nov 2017 – Dec 2018 · 1 yr 1 mo · On-site

o SQL Quarantine: Created a mechanism to automatically quarantine SQL plans/statements having a history of violating Oracle Resource Manager limits.
o SQL Test Case Builder: Improved functionality by exposing metadata via SQL system views and implementing file compression for exported cases.
o Oracle Sharding: Captured shard usage in SQLs and exposed it in SQL system views, and enhanced plan display packages to show shard-level execution details.
o Supported continuous improvement and maintenance of RDBMS Optimizer, SQL Sharding, SQL Test Case Builder, and SQL Quarantine.

SQL QuarantineSQL Test Case BuilderSharding

Member Of Technical Staff

Jul 2016 – Oct 2017 · 1 yr 3 mos · On-site

o Oracle Sharding: Advanced Oracle Sharding capabilities by integrating shard usage tracking in SQL system views and augmenting plan display packages with shard-level execution insights.
o Actively contributed to the Optimizer and Sharding.

ShardingOptimizer

University of illinois at urbana-champaign

Visiting Scholar / Undergraduate Project

May 2015 – Nov 2015 · 6 mos · Urbana-Champaign, Illinois, United States and Kanpur, Uttar Pradesh, India · Hybrid

Worked as the part of Blue Waters Petascale Application Improvement and Discovery(PAID) program.
Profiled MPI-based C/Fortran applications to collect communication and performance data for point-to-point and collective operations. Analyzed patterns to evaluate the impact of network interference..
Researched polynomial-time approximation schemes for QAPs and heuristics to map application communication graphs onto physical network topologies, reducing communication overhead.

MPICFortranPerformance Analysis

Indian institute of technology, kanpur

Undergraduate Student

Jul 2012 – Jun 2016 · 3 yrs 11 mos · Kanpur Nagar, Uttar Pradesh, India · On-site

Graduated with a Cumulative Performance Index (CPI) of 9.3/10.0.
G-Skew branch predictors: Studied the performance of G-Skew branch predictors with FNBT, bimodal, SAg, GAg, gshare predictors along with their hybrids. Predictors were simulated using PIN on Spec 2006 benchmarks.
File sharing system: Developed a locality aware peer to peer file sharing system with caching. TCP was used for tracker-client communication and UDP for peer to peer communication.
Pascal compiler: Designed a basic pascal compiler, and enhanced it to support cache efficient code generation(conditional statements), constant folding, I/O library, N-dimensional arrays.

CFortranMPI