Design Engineer I
India
* Worked with DSP IP team for verification of multiple Audio Codecs, Radar-Lidar Libraries and Neural Network SDK* Automation of testing infrastructure using Python, Jenkins, Linux Shell Scripts, C++, Matlab, LSF for significantly minimising validation time* Prompt verification of customer & internal bugs for multiple products involving debugging to identify issues & collaborative cross-team effort for quick & efficient fixes1. NDSP Libraries (Optimised math & signal processing library)- * Functional breakdown of legacy C++ library using OOPS for automated parallel execution on LSF * Reduced verification time from hours down to minutes by designing a framework using C/C++, Matlab, and Python2. UTF- * Development, debugging and bug fixing for UTF, a multi-threaded end-to-end automated framework for verification and release process using Django, SQL, Python, Linux Shell Scripting, Jenkins, JIRA, Perforce, etc * Designed a scalable and durable database structure to facilitate usage of Machine Learning in test case generation and execution3. NNE 110 SDK and NNLib- * Creation of neural networks as inputs and regression setup for Neural Network Engine 110 SDK * Sequence files for Neural Network library using Python and Matlab (NNLib has optimized NN operators used in speech and audio workloads)4. Other- * Created multiple configurations of DSP cores using Xtensa Processor Generator * Performed quality checks for multiple products like Code Coverage (gcov, xt-cov), static code analysis (MISRA-C), build checks with Xtensa LLVM C/C++ compiler (xt-clang), buffer alignment, ping-pong, etc * Designed tests, functionality testing, performance analysis (MIPS), and debugging for Libraries and Codecs under RLC and Audio * Setup weekly regressions of Radar-Lidar Libraries on Xtensa Toolchain (Recent Good Build) * Integrated execution of a Windows tool through Linux server scripts using TCP/IP* One of the youngest participants in the Cadence Women Conference 2023
