Hi, I’m
Yashodhan Phatak
Senior Staff System Software Engineer
San Francisco,CA
Summary
Results-driven engineer with extensive experience in designing and developing advanced distributed systems and infrastructure. Proven expertise in systems architecture and software development, with a strong track record of delivering scalable, high-performance solutions. Skilled in research and development of machine learning models, leveraging deep technical knowledge and hands-on experience to drive innovation and solve complex problems.
Overview
17
years of professional experience
Work History
UC Berkeley
Machine Learning Engineer (ML Professional Certification)
06.2024 - Current
Job overview
- Researched and developed numerous machine Learning high-performing models targeting real-world problems
- Utilizing techniques involving exploration data analysis and model selection, training, validation, fine tuning
- Extensive experience with supervised learning models such as Linear regression, Logistic Regression, Decision Trees
- Excellent understanding of Deep Learning (DL) models such Deep Neural networks, CNNs, LSTMs, RNNs, Transformers
- Unsupervised models: PCA, SVD, K-Means Clustering
- Conducted rigorous model validation and performance assessments to achieve robustness of ML predictions.
- Utilized model Optimization techniques such as loss function, regularization, gradient descent, bias-variance, metrics
Intel Corp.
Senior Staff System Software Engineer
06.2008 - 05.2024
Job overview
- Company Overview: DATA CENTER and AI (DCAI)
- Research & development of distributed system for 5G Network to enable it as Infrastructure-as-a-Service on the Datacenter, Network platforms.
- Large-scale distributed infrastructure design and development with cloud native technologies
- Development of scalable containerized stateless microservices for main network components like AMF, SMF, UPF.
- Orchestration of network function micro-services with Kubernetes for optimal utilization of available servers
- Load balancing on the microservices with worker-thread pool for servers for robustness
- RESTful API based interfaces between main network components for scalability
- DATA CENTER and AI (DCAI)
Intel Corp.
Staff Embedded Software Engineer
Job overview
- Company Overview: Client Compute PC Group (CCG)
- Research & development of system algorithms for traffic arbitration(coexistence) between Bluetooth and Wi-Fi to optimally share the resources, for multiple generations of connectivity targeting Intel PC platforms.
- Developed Coexistence system firmware which is shipping on billions of PCs providing best-in-class user experience on Wi-Fi and Bluetooth connectivity.
- Real-time interface(N-wire), non-real time interface (Mailbox) and messaging protocols between Bluetooth SoC and collocated Wi-Fi SoC; architecture/design/development.
- Developed great insights in CPU architecture while porting software stack from ARM SoC to ARC SoC.
- Utilized compiler toolchain to rewrite optimized code assembly, inline assembly and with compiler specific changes in C code for new processor ISA.
- Platforms bring up, with ported code on new SoC Rewrite for software optimization.
- Released optimized patch software built using assembly programming.
- Introduced robust coding practice by using defensive programming techniques like assert macros, checks for buffer overflow, null pointer deference and enforcing it in build environment with Klocwork.
- Performed I2C interface protocol emulation using bit banging technique. Used this facility for debug.
- Supported many PC OEM/ODMs with critical field issues by debugging using traces, JTAG, OTA sniffer logs.
- As a technical leader, he provided mentoring to junior and new team members.
- Client Compute PC Group (CCG)
- Technologies used: ARM CPU, Hardware SoC architecture, Software optimization, low-level firmware driver, C, RTOS(Real-time os), JTAG, interfaces - (UART, USB, I2C, Mailbox), LA, Klockwork
Skills
large-scale Distributed system design
scalable platform-infrastructure
Stateless Microservices
RESTful API
OpenAPI standard
RedisDB
containerization
orchestration
Dockers
Kubernetes
C
Python programming
Numpy
Pandas
Sklearn
Scipy
Nltk
Seaborn
machine learning
deep learning models
TensorFlow
PyTorch
low-level operating systems concepts
multi-threading
memory management
networking
performance
scale
C
sellscripting
Assembly
ARM CPU architecture
System Performance optimization
Software Performance optimization
Technical leadership
coaching
mentoring
Education
University of North Carolina
Master of Science (M.S.) from Computer Engineering
University Overview
- Research topic: Distributed computing
- Extracting key profiling Metrics from the network using unique microservice API
- Test framework for validating the platform infrastructure
- Servers (SCTP, HTTP, UDP based) and wrapper library exposing uniform APIs for communication
- Mentored junior members to bring them to speed on projects, resulting in quicker completion milestones with this IaaS is shipping with hundreds of thousands of Network Platforms
University of Pune
Bachelor of Electrical and Electronics engineering
Timeline
Machine Learning Engineer (ML Professional Certification)
UC Berkeley
06.2024 - Current
Senior Staff System Software Engineer
Intel Corp.
06.2008 - 05.2024
Staff Embedded Software Engineer
Intel Corp.
University of Pune
Bachelor of Electrical and Electronics engineering
University of North Carolina
Master of Science (M.S.) from Computer Engineering
06.2024
Publications
Distributed systems and access using Cluster of FPGAs, IEEE Xplore/Print, 978-07695-3307-0