Vivek K. Email & Phone Number

California, United States

Mountain View, California, United States

San Francisco Bay Area

• Lead software development for delivering best in class AI inference performance with MemryX AI accelerator.• Deliver End to End Edge AI software solution for customer engagements in multiple verticals and platforms.• Develop next gen product requirement by identifying customer requirement, market trends and drive ecosystem partnerships.

San Francisco Bay Area

• Lead AI research group to implement, customize and optimize SOTA AI models on Vision, NLP and GAN use cases• Lead cross-functional team to deliver higher throughput & lower latency with periodic releases on TensorFlow for Intel Client XPUs• Establish on device learning development and KPI measurement infrastructure for incremental model updates for LLMs and GANs• Define requirements and establish KPIs on next generation hardware and software products on Client & Edge platforms and use cases • Presented at CVPR'22 workshop on AI optimizations for Intel Platforms• Collaborate with product management, marketing and key customers to define requirements and establish KPIs on next generation hardware and software products on Client platforms and use cases• Lead cross functional team to characterize concurrent AI workloads on CPU, GPU & VPU and derive H/W, S/W optimization for current and future usescases on nextGen platforms• Architect, plan and deliver deep learning software stacks for Edge AI devices integrated into 25+ products spanning industries including retail, smart cities, healthcare, and manufacturing• Led cross-functional teams to improve model performance using techniques of model pruning, quantization, and compression techniques• Design and deliver the software stack for end-to-end inference pipeline for Analytics viz. Object detection, Image Classification, Segmentation, Human pose, face detection/recognition involving AI, Orchestration, and Manageability• Initiated and led R&D team to enable NLP use cases at Client Edge devices with Transformer models, develop Neural Architecture Search (NAS) models to optimize performance and accuracy by 30%• Collaborated with ISVs and solution providers to manage reference designs and develop ecosystem strategy• Define and develop Firmware and application Pipeline for use cases of Smart Camera using traditional CV and DL video analytics

San Francisco Bay Area

• System Software Architect for IoT Video Development Kit based on Embedded Linux OS. Worked on integration of Camera sensor, Wireless module and proprietary AE algorithm. • Developed 3A camera algorithms for 2 gens of Camera ISPs and optimized on firmware to productize for more than 25 customers.• Spearheaded the design and development of kernel drivers, middleware, and applications on Android-based reference platforms to enhance sensor capabilities• Defined ASIL requirements, computation, and cost complexity for use-cases and algorithms for rearview and in-cabin smart camera• Designed the control and data flow between different processing modules for next-generation Automotive ISP Firmware• Integrated all-new PDAF & OIS sensors on Qualcomm platform modified their AF framework and implemented in 10+ products

Bengaluru Area, India

• Designed solutions for Aptina’s Image sensors and Image processors and spearheaded product integrations with customers and technology partners• Integrated Aptina Sensors/Image Processor on Qualcomm platforms (Android OS) with 4K video capture and playback which was showcased in MWC and CES conferences• Contributed multiple patches in opensource Tactical Driver Development on supporting YUV and RGB data processing on ISP submodule in V4L2 media-controller for NXP and OMAP platforms• Implemented OTPM Framework to read sensor data related to calibration and tunings like LSC, AWB, and AF for continuous integration with Image Pipeline and applied to all future products• Experience of Board and Sensor bring up on many platforms; Understanding of hardware design & schematics; Extensive debugging on camera interface lines I2c and CSI-2

Bengaluru, Karnataka, India

• Envisioned, and developed Eklavya, an educational content development and distribution platform• Motivation to provide an unique interactive learning experience to high-school students at value-based pricing• Top 20 finalist in Power of Ideas, a pan-India startup competition

Taipei, Taiwan

• Maintained BSP for Embedded Linux OS and developed peripheral device drivers viz. Wireless, SPI and Touchscreen interfaces• Implemented SLAM and developed Computer vision algorithms for EeeBot motion planning and movement• Developed user-friendly applications based on QT Framework for EBook Reader and EeeTablet

Leuven, Belgium

Worked on data mining in the Department of Computational Neuroscience, Neurofysiologie department. • Developed end to end application for signal analysis of very large data files to extract into meaningful text files for need-based interpretation of different attributes using K-means and PCA which significantly helped medical researchers at the laboratory to cut down their analysis time by more than 50%

Ai Graduate Program

* CS221- Artificial Intelligence: Principles and Techniques: Machine learning, Search, Game playing, Markov decision processes, Constraint.

Master Of Technology - Mtech, Electrical Engineering

Bachelor Of Technology - Btech, Electrical Engineering