Mayank Bhaskar Email & Phone Number

Lucknow, Up, In

Lucknow, Uttar Pradesh, India

I am one of the core organizers of the Lucknow AI Labs where I help the team to select papers, topics and also mentor them on what to present in the session 📜I worked on DistAya: From the Expeditions Aya, Cohere For AI group that includes dabbling with pruning, sparsity & knowledge distillation of Multilingual LLMs to compress them to smaller models that can be run with large batch sizes on the cloud, mobile & edge devices with higher throughput, lower latency, and greater inference… Show more I am one of the core organizers of the Lucknow AI Labs where I help the team to select papers, topics and also mentor them on what to present in the session 📜I worked on DistAya: From the Expeditions Aya, Cohere For AI group that includes dabbling with pruning, sparsity & knowledge distillation of Multilingual LLMs to compress them to smaller models that can be run with large batch sizes on the cloud, mobile & edge devices with higher throughput, lower latency, and greater inference speeds 🛠️Research proposal on Deep Banach Space Kernels for the CoSubmitting Summer @ ICLR 2022 whose goal was to define a representer theorem for Deep/Compositional Reproducing Kernel Banach spaces and potentially derive a deep kernel that reproduces the Banach space whose solution set should contain one of the well-known Neural Network architectures (e.g. ReLU Nets) 💬On top of that, my team finished 3rd on the FLAX/JAX community event organized by HuggingFace, Fine-tuning CLIP with Remote Sensing (Satellite) images and captions 🖼️Having attended top ML Conferences like ICLR, ICML, CVPR & NeurIPS remotely, I have the insights, and experience to help researchers bounce ideas off and work towards achieving our combined goals 🌎Moreover, I am involved in the Daily & Weekly Paper Discussions on Yannic Kilcher's Discord server to keep up with research & their applied use cases in the industry 🏭Learning advanced CUDA & Triton for stepping towards native GPU programming and making distributed training and inference go vroom! Part of the CUDA-Mode Discord server to achieve applied AI's high-performance goals 🚀 Show less

St Louis, Missouri, United States

I contributed to open-source projects, led, moderated, and presented in sessions about Generative AI, Diffusion, MLOps, and Reinforcement Learning. Also, I am involved in Kaggle and a small group of us meet every week on Saturdays, where I won a silver & a couple of bronze medals participating in 17 competitionsWorked on RAGs and made open-source contributions to the TWiML-RAG project where I helped develop pipelines for transcription & diarization of podcasts (via STT) using OpenAI's… Show more I contributed to open-source projects, led, moderated, and presented in sessions about Generative AI, Diffusion, MLOps, and Reinforcement Learning. Also, I am involved in Kaggle and a small group of us meet every week on Saturdays, where I won a silver & a couple of bronze medals participating in 17 competitionsWorked on RAGs and made open-source contributions to the TWiML-RAG project where I helped develop pipelines for transcription & diarization of podcasts (via STT) using OpenAI's WhisperI also help moderate Generative-AI meetups every Friday where we connect, share, and discuss topics on Generative-AI (Langchain, Agents, etc), RAGs, LLMs, Computer Vision & Multimodal Models (VLMs), delve into ML techniques & algorithms, breakthrough research & their use-cases in the real world, and more⚡️ Competition - Meta & Wandb AI Hacker Cup⚡️ Competition - Google AI Hackathon⚡️ Competition - HuggingFace's HackathonSlides & YouTube playlists of some talks and sessions:📌 Building Machine Learning Pipelines: Automating Model Life Cycles with TensorFlow by Hannes Hapke, Catherine Nelson - Book study group (+ Mash Zahid + Wenchang)https://bit.ly/3zl4hZK📌 Hands-on Machine Learning with Scikit-Learn, Keras and TensorFlow, by Aurélien Géron - Book study group (+ Joseph Catanzarite + Mash Zahid)https://bit.ly/3z8tSFf📌 CS224W: Machine Learning with Graphs 2021 Study Grouphttps://bit.ly/3ZoiYpi📌 Reinforcement Learning: An Introduction, Book by Andrew Barto and Richard S. Sutton - Book study group (+ Arto)https://bit.ly/4eoo6hr📌 DEEP LEARNING (with PyTorch) course by NYU (2020)https://bit.ly/4e5v6QE📌 Natural Language Processing with Transformers by Leandro von Werra, Lewis Tunstall, and Thomas Wolf - BookClubhttps://bit.ly/3Tox6uP📌 Practical DL for Coders (Part 2) - Study Grouphttps://bit.ly/3zk81dM📌 Deep Unsupervised Learning (Berkeley) 2021 by Pieter Abbeel et. al.https://bit.ly/4epDj1I📌 Kaggle Meetuphttps://bit.ly/3TtowLC Show less

Tokyo, Japan

I contributed, led, and moderated remote sessions about ML Maths, Machine Learning, FastAI, Reinforcement Learning & Generative AI. Slides & playlists of some talks and sessions:📌 NeurIPS-21 Research Paper Discussion (+ Mrityunjay Bhardwaj)https://bit.ly/3a2eYQS📌 Reinforcement Learning: An Introduction, Book by Andrew Barto and Richard S. Sutton Cohort (+ Raymond Cheu + Mrityunjay Bhardwaj + Anugraha Sinha + Suzana Ilić)https://bit.ly/3dULq94📌 Mathematics for… Show more I contributed, led, and moderated remote sessions about ML Maths, Machine Learning, FastAI, Reinforcement Learning & Generative AI. Slides & playlists of some talks and sessions:📌 NeurIPS-21 Research Paper Discussion (+ Mrityunjay Bhardwaj)https://bit.ly/3a2eYQS📌 Reinforcement Learning: An Introduction, Book by Andrew Barto and Richard S. Sutton Cohort (+ Raymond Cheu + Mrityunjay Bhardwaj + Anugraha Sinha + Suzana Ilić)https://bit.ly/3dULq94📌 Mathematics for Machine Learning (+ Emil Vatai + Mrityunjay Bhardwaj)https://www.youtube.com/playlist?list=PLaPdEEY26UXygpV-Cxch8Xkpl7IbFKBvK📌 FastAI Cohort (+ Sanyam Bhutani)https://course.fast.ai/📌 Generative Deep Learning by David Foster, Book reading and project cohort (+ Suzana Ilić + Emil Vatai + Mrityunjay Bhardwaj + Anugraha Sinha)📌 High Performance Python, 2nd Edition, Book Cohort (+ Sanyam Bhutani + Mani Sarkar)📌 Python for Data Analysis, 3rd Edition, Book Cohort (+ Sanyam Bhutani + Mani Sarkar)📌 Deep Learning with Python Study Series based on the book “Deep Learning with Python” by François Chollet (+ Dimitris Katsios)https://www.youtube.com/playlist?list=PLaPdEEY26UXyFIuJaDvb7nSlUUOSdBKpP📌 MLT Artificial Intelligence Talks from the academia & industry experts (+ Suzana Ilić)https://www.youtube.com/playlist?list=PLaPdEEY26UXzDgr0uN1OUyLdeTwd477l2📌 CNN Architectures - implementations (+ Dimitris Katsios)https://www.youtube.com/playlist?list=PLaPdEEY26UXyE3UchW0C742xh542yh0yI Show less

Indore, Madhya Pradesh, India

My role evolved, initially I worked as an Engineer and subsequently took on a Consultant roleI worked on training computer vision (CV) models via style transfer in a Lua-based environment and efficiently pruning, distilling, and quantizing them to an Apple's iOS on-device ecosystem of models. I was denied a US VISA when I was chosen to go for Apple's WWDC 📲Moreover, I led an augmented reality 3D animation endeavor, focussing on Apple's ARKit marker & location tags using video… Show more My role evolved, initially I worked as an Engineer and subsequently took on a Consultant roleI worked on training computer vision (CV) models via style transfer in a Lua-based environment and efficiently pruning, distilling, and quantizing them to an Apple's iOS on-device ecosystem of models. I was denied a US VISA when I was chosen to go for Apple's WWDC 📲Moreover, I led an augmented reality 3D animation endeavor, focussing on Apple's ARKit marker & location tags using video data coupled with LiDAR. Plus, using the LiDAR scanner to create a polygonal model of the physical environment 📸Additionally, I was involved in applied research of 3D pose estimation & tracking along with mesh-based human modeling. Also, I got my hands dirty with the practical applications of hand, arm, shoulder, and facial features tracking via custom components of Google's MediaPipe amongst other opensource libraries 🔍Furthermore, I have delved into developed applications based Nvidia's DeepStream SDK streaming analytics system (Deep Learning) - writing custom kernels, training, fine-tuning, and then providing inference optimization on vision models dealing with video data that was detecting cars from a full or empty parking spot along with streaming live data from a traffic intersection to detect number plates and other features 🚗Moreover, I was involved with Intel® RealSense™ Depth Camera D435i, developing applications for different use cases including a 3D mesh-based human modeling system 📸In addition to that, I have worked in many reinforcement learning simulators - Amazon's DeepRacer, Microsoft's AirSim, Carla & Autoware's ROS2 system, and many open-source robotics ecosystems including DonkeyCar, Nvidia's Isaac, amongst others 🤖Utilized machine learning to categorize EEG signals for a wearable brain device focused on neurofeedback 🧠Led Innovation Tuesday meetings where a group of curious minds discussed groundbreaking research and its applications 🔬 Show less

Lucknow, Uttar Pradesh, India

Research & Development on how Google's Android is bringing in a revolution, i.e. how the existing systems are being integrated with a new set of devices to leverage artificial intelligence and bring transformation coupled with innovation in the enterprise.I was involved in grasping the architecture of the Android platform and its underlying framework. Using the Android Open Source Project (ASOP), I built a functioning version of Android from the ground up. This involved setting up and… Show more Research & Development on how Google's Android is bringing in a revolution, i.e. how the existing systems are being integrated with a new set of devices to leverage artificial intelligence and bring transformation coupled with innovation in the enterprise.I was involved in grasping the architecture of the Android platform and its underlying framework. Using the Android Open Source Project (ASOP), I built a functioning version of Android from the ground up. This involved setting up and using many different tools to assemble a system image which was complete enough to include the Android kernel, as well as other essential parts like the OS, binaries, and all of the other features needed for a device to run Android. This was a long and enlightening project, during which I gained a great deal of experience with both Android and general OS assembly.I used the Gradle build system to develop REST-based Web Services and some sample apps for consuming data. In addition to that, I played around with Virtual Reality (VR) - via Google Cardboard SDK to build a couple of apps from scratch, getting 3D assets from the open-source Unity and Blender stores. Show less

Lucknow, Uttar Pradesh, India

Broadleaf Commerce provides B2B and B2C eCommerce platform solutions to simplify the complexities of multi-channel commerce and digital experience management. The objective was to use the community edition of the eCommerce solution and add B2C features like a Shopping Cart, Offer & Promotion, Order Management System, CMS, etc. to form a product for a small-medium sized business.

Lucknow, Uttar Pradesh, India

Data Migration [Product], Center of Excellence Automation Lab, AssuranceDevelopment of a data migration product, which focused on the migration of data from one of the products to another.Product Development Responsibilities:- Feasibility analysis of HP ALM, ServiceNow & Atlassian JIRA- Architecting of a batch scheduler- Consuming and Producing REST-based Java Web Services

Lucknow, Uttar Pradesh, India

R&D in Android, for feasibility analysis of its architecture, hardware & software requirements to integrate with our existing set of enterprise-level applications in the CoE Lab.Feasibility Analysis Responsibilities:- Android Architecture- Developing apps like parsers, contacts, amongst others to get a holistic understanding of the underlying framework- Android Studio, Gradle- Development of REST-based Java Web Services for producing and consuming data

Lucknow, Uttar Pradesh, India

E-Assure [Product], Center of Excellence Automation Lab, AssuranceDevelopment of product E-Assure, which included modules for Automation, i.e. processing of test cases from the HP ALM tool, Analytics, i.e. ability to convert the data into meaningful information, and User Management, i.e. portal for creating and managing different users based on their roles & responsibilities.Product Development Responsibilities:- Actively participated in designing of logical data modeling… Show more E-Assure [Product], Center of Excellence Automation Lab, AssuranceDevelopment of product E-Assure, which included modules for Automation, i.e. processing of test cases from the HP ALM tool, Analytics, i.e. ability to convert the data into meaningful information, and User Management, i.e. portal for creating and managing different users based on their roles & responsibilities.Product Development Responsibilities:- Actively participated in designing of logical data modeling and developing of physical data modeling.- Development of the base architecture for the application, i.e. integration of Springs, Hibernate, Connection Pooling mechanism, and Jasig CAS Single Sign-On.- I was also involved in the design and development of a web interface for the User Management component with jQuery, Ajax, and jqWidgets.- Designed the overall architecture of the Scheduler and License mechanism using various design patterns.- Worked on various Tomcat & Apache Commons Daemon deployment activities- Usage of Maven as a build tool management- Wrote Springs JUnit test cases Show less

Noida, Uttar Pradesh, India

ESS-IVR, Walgreens[1] Worked as a Java developer in the Walgreens, ESS-IVR project which reflected the Interactive Voice Response, using open-source technologies like VXML. I developed some basic modules for the IVR and also collaborated with developing a cache for faster retrieval of information from the database.[2] Was a part of maintaining a reporting tool, which displayed information from multiple Walgreens pharmacy stores, taking data from the Avaya system, into Oracle… Show more ESS-IVR, Walgreens[1] Worked as a Java developer in the Walgreens, ESS-IVR project which reflected the Interactive Voice Response, using open-source technologies like VXML. I developed some basic modules for the IVR and also collaborated with developing a cache for faster retrieval of information from the database.[2] Was a part of maintaining a reporting tool, which displayed information from multiple Walgreens pharmacy stores, taking data from the Avaya system, into Oracle RDBMS, running a nightly batch.Responsibilities:- As Jr. Developer, I was actively involved in architecting various Business Layer and Data Management components of this multi-tiered Voice XML(VXML) based system over b0unce framework, custom JSP tag-libs, etc.- Responsible for developing and maintaining all the entity and session beans.- Developed various Customer flows. Developed various server-side components using J2EE.- Maintained the distributed batch system for tackling data from the Avaya system- WebSphere configuration for mailing Show less

Hyderabad, Telangana, India

Telecom Inventory Management System, Initial Learning Program(ILP)The objective was to develop a Java EE based web software that facilitates wholesalers to sell their products to retailers in the market.Responsibilities:- First, was the User Interface module, which was about designing the interface for the retailer.- Second, was Stock Management module, which dealt with storing the information about products & their accessories, creating offers, replacing defective… Show more Telecom Inventory Management System, Initial Learning Program(ILP)The objective was to develop a Java EE based web software that facilitates wholesalers to sell their products to retailers in the market.Responsibilities:- First, was the User Interface module, which was about designing the interface for the retailer.- Second, was Stock Management module, which dealt with storing the information about products & their accessories, creating offers, replacing defective products, updating the quantity each time a product was ordered, etc.- Third, was the Order Management module, which facilitated the retailer to place & cancel orders Show less

Pune, Maharashtra, India

Interned on XHTML, JavaScript, jQuery, PHP backend, and CMS(Content Management System) 🖥Moreover, learned Objective-C and contributed to the iOS development of an app 📱

Bachelor Of Technology - Btech, Computer Science & Engineering, 70.82%

Activities and Societies: Worked on multiple Web and Enterprise applications with Java, Servlets, and Struts based projects. Deployed.

Higher Secondary - Central Board Of Secondary Education (Cbse), Physics, Chemistry, Math, Computer Science, English, 83.8%

Activities and Societies: I worked on multiple C++ projects with data structures and developed many games using the display library.

High School - Central Board Of Secondary Education (Cbse), English, Hindi, Mathematics, Science, Social Sciences, Art, And Physical Education, 64.4%

Activities and Societies: I worked on web-based projects using HTML, Cascading Style Sheets (CSS) & JavaScript on Adobe Dreamweaver.