About Me

As a seasoned machine learning engineer with over 4 years of experience, I embody a fluid-like adaptability that seamlessly bridges gaps within a team. Whether leading or collaborating, my focus is on identifying and filling the spaces where innovation thrives, bringing both leadership insights and a collaborative spirit to drive positive change in the dynamic world of technology.

Components from my ML Experience

Languages

o   Python
o   C++
o   R
o   Matlab
o   MySQL

Operating Systems

o   Linux/Ubuntu
o   Windows
o   ROS

Libraries

o   Pytorch
o   Tensorflow
o   NumPy
o   Pandas
o   Keras
o   SciPy
o   Matplotlib
o   Scikit-Learn
o   Mediapipe

Algorithms

o   XGBoost
o   Logistic Regression
o   Multi-Armed Bandit
o   YOLOv3
o   R-CNN
o   Faster R-CNN
o   SSD
o   ResNet50
o   VGG16
o   BERT
o   RoBERTa
o   NLTK
o   spaCy
o   Diffusion
o   Segment-Anything
o   Detectron2

Tools

o   Unity
o   Google Dialogflow
o   A1111
o   Docker
o   Git

Techniques

o CI/CD
o Stacking

Testing

o A/B testing

Cloud Computing

o   AWS
o   Azure
o   Auto-scaled Instance

My contribution

Machine Learning Engineer

Crafty Apes VFX

Created various CAT nodes as part of the Applied R&D team to reduce the time spent by VFX artists by 50%.
Led a team of 2 Machine Learning engineers to develop models that semi-automate the process of beautification(such as de-aging, and removing blemishes) and compositing(such as creating frames based on textual descriptions, removing/adding objects to the frame, and removing shadows). This improved the quality of the end results by 70%.
Worked with 4 different VFX artists to come up with efficient workflows for each use case.
Developed Landmark detection and object tracking models to help with the editing process

Cognitive Analysis and Machine Learning Engineer

Microsoft (Mindtree)

Analyzed the specific needs of each customer to provide them with suggestions on Azure solutions that would best fit them.
Ensured the robustness and accuracy of Azure models through meticulous attention to detail in the development process of each upgrade.
Monitored training of Speech and Language models to optimize cost and time
Conducted in-depth analysis to discern and address discrepancies between the anticipated and actual outcomes of Bing API

NLP intern

Glendor

Researched and tested the performance of BERT, RoBERTa, NLTK, and spaCy for each PII entity present in medical records.
Developed a NER model using the comparative analysis

Jr.Computer Vision Engineer Intern

Sensagrate

Created a road-assistance model incorporating object detection, object tracking, and path prediction to improve road safety of drivers. I was the solo contributor of this model.
Created a front-end user application using PyQt
Conducted research for optimized solutions using sensor-fusion

Research Aide

Arizona State University

Created a Simulation Environment on unity and activated communication between the Microsoft HoloLens-Robot and Microsoft HoloLens-simulation through ROS.
Assisted through the research and development of a prototype of navigation of a robot using HoloLens.

Software Developer (Chatbot Training)

Verizon (Experis IT Solutions Pvt LTD)

Managed user-end dialogues to chatbot using Google Dialogflow
Cleaned the data of the existing conversations to train the chatbot and improve its accuracy to 80%
Trained the bot for keyword-based and context-based understanding

Data Scientist

Smart Kredit Tech Pvt LTD

Completed a design of a predictive analysis of unclean data for an automobile insurance company without knowledge transfer within a period of 2 weeks and improved the accuracy from 50% to 74.2%. Led a team of 3 data analysts to accomplish the same.
Designed and developed a machine learning model to analyze and direct digital ads on the basis of interests and demographics. In the process of completion of the application, filled in the gaps of the team for backend support and UI/UX testing and debugging.