Ankur Narwar

Software Engineer

Hyderabad, Telangana, India10 yrs 5 mos experience

Most Likely To SwitchAI ML Practitioner

Key Highlights

Expert in optimizing deep learning frameworks.
Strong background in system programming and inter-processor communication.
Proven ability to enhance software performance significantly.

Stackforce AI infers this person is a Machine Learning and Software Development expert in the Embedded Systems and AI/ML industry.

Contact

Skills

Core Skills

Machine LearningSoftware DevelopmentSystem Programming

Other Skills

AI software TeamSNPE frameworkhardware accelerated inferencesDepthwise ConvolutionMulti-class NMSpeak memory optimizationSIMD parallelismQA frameworksperformance analysisdebuggingPerformance OptimizationMPROCinter-processor communicationWindows Desktop/Phonetransport plugins

About

I am Ankur, an enthusiastic Computer Science graduate and have completed M.Tech from NIT Trichy. I am interested in Machine Learning, Algorithm Design, Problem Solving, System Programming and Networking. My objective is to pursue a challenging career that gives me scope to enhance my knowledge, skills in the computing and research field with sheer determination, dedication and hard work.

Experience

10 yrs 5 mos

Total Experience

5 yrs 2 mos

Average Tenure

9 yrs 11 mos

Current Experience

Qualcomm

5 roles

Staff Engineer

Promoted

Dec 2024 – Present · 1 yr 5 mos

Senior Lead Software Engineer

Nov 2021 – Nov 2024 · 3 yrs

Senior Engineer

Promoted

Dec 2018 – Oct 2021 · 2 yrs 10 mos

Member of the AI software Team. Working on the SNPE framework which enables efficient inferencing of various deep neural networks on Qualcomm SOCs
Optimizations for hardware accelerated inferences on CPU Runtime
Optimized performance of multiple layers like Depthwise Convolution and Multi-class NMS by a huge margin
Optimized peak memory usage of SNPE to a great extent
Developed a module to provide high performance SIMD parallelism on multiple CPU cores
Developed and optimized QA frameworks to increase quality checks of the product
Analysis and debugging of performance issues of the neural networks

AI software TeamSNPE frameworkhardware accelerated inferencesDepthwise ConvolutionMulti-class NMSpeak memory optimization+6

Machine Learning Software Engineer

May 2018 – Nov 2018 · 6 mos

Member of Qualcomm's first Machine Learning team in India that works on Snapdragon Neural Processing Engine (SNPE) - Qualcomm Snapdragon software accelerated runtime for the execution of Deep Neural Networks.

Engineer

Feb 2016 – Apr 2018 · 2 yrs 2 mos

Engineer in MPROC (Multi-Processor Communication) team under Core BSP (Board Support Package) dept.
MPROC provides high performance inter-processor communication stack encompassing all layers
from session to link layer used for messaging within the Qualcomm Snapdragon SoC
Leads the Windows Desktop/Phone department corresponding to MPROC
Developed multiple transport plugins to communicate with newly added processors on SoC
Drove the efforts impacting our modules due to Hibernate feature requirement on new chipsets
Developed multiple debugging scripts in python to reduce the debugging time by 50%
Design, development and support of multiple layers of drivers for inter-processor communication on Snapdragon SoC
Pre-silicon planning and development to further expedite feature delivery