Help & Supportcareers@gravityer.com
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Leading Product & AI @ Realbotix |
I bring a unique combination of AI expertise, Mechanical Engineering experience and leadership in product execution, placing me at the intersection of robotics, intelligent systems, and real-world business impact. I currently lead the Software and AI team for a robotics company, where my work focuses on designing scalable AI architectures for embodied systems, including LLM integration, memory, vision, RAG pipelines, conversational intelligence, and edge/cloud deployment. My focus is on building scalable AI systems, leading high-performing teams, and translating emerging technologies into practical business value. Please go through my profile to know more about me!
Indiana University Bloomington
Master's degree, Data Science
August 1, 2022 – August 1, 2024
Manipal Institute of Technology
Bachelor of Technology - BTech, Mechanical Engineering
January 1, 2017 – January 1, 2021
Realbotix (TSXV: XBOT | OTCQB: XBOTF | FSE: 76M0.F)
Head of AI
July 1, 2025 – Present
Las Vegas, Nevada, United States
Realbotix (TSXV: XBOT | OTCQB: XBOTF | FSE: 76M0.F)
AI-Architect
September 1, 2024 – July 1, 2025
Las Vegas, Nevada, United States
Realbotix (TSXV: XBOT | OTCQB: XBOTF | FSE: 76M0.F)
Generative AI Engineer
July 1, 2024 – September 1, 2024
Las Vegas, Nevada, United States
IU Health Bloomington Hospital
Gen AI solution developer
March 1, 2024 – June 1, 2024
Bloomington, Indiana, United States
Indiana University Bloomington
Associate Instructor
January 1, 2024 – May 1, 2024
Bloomington, Indiana, United States · On-site
Indiana University Bloomington
Graduate Teaching Assistant
August 1, 2023 – December 1, 2023
Bloomington, Indiana, United States · On-site
Four Colors Technology
Data Science Intern (LLM modeling)
May 1, 2023 – July 1, 2023
Boston, Massachusetts, United States · Remote
DataOrc
Data Scientist (Machine Learning)
August 1, 2021 – June 1, 2022
Pune, Maharashtra, India · Remote
IBM
Data Scientist (Financial Modeling)
September 1, 2020 – July 1, 2021
Pune · Remote
News Maker's Network (Natural Language Processing)
November 1, 2023 – November 1, 2023
Skills : Vector Databases, Transformers, Named Entity Recognition with BERT Transformer, Spacy Transformer, Sentiment Analysis, Network Models, Flask, HTML, NLTK As we fast approach the US elections, it is hard to analyze the public sentiment around an entity. As humans, we can get quickly opinionated on a 'window-of-observation' which we see on the news. Aim : We aimed to create a tool that users can use to quickly visualize the co-mentions for any named entity (for instance, Donald Trump) in the news and see their relative sentiment, as well as overall sentiment values, with 10 things they are most associated with. Dataset : We used the CNN News Articles containing over 9,000 articles from CNN's website. This dataset is downloadable from Kaggle here: https://www.kaggle.com/datasets/hadasu92/cnn-articles Overview of the Steps : - Data Preprocessing : News articles had to be cleaned using various NLP techniques like stemming, tokenization, lemmatization, etc. - NER : We then explored multiple models from BERT Transformers, Spacy's models (small,large,transformers), processing the data as required by each. - After creating comparitive models, we used the Spacy's Transformer model in our pipeline for NER. Advantages of using Spacy : We can use transformer embeddings like BERT directly in spaCy which enabled us to quickly compare and train the model for our needs. We used non context based approach (word vectors) as well as context based(Transformers). Functionalities: - Our app creates treats the given entity(person) as a central node and creates a network graph of top 30 people that our given entity has appeared the most in the articles with. - The thickness of the edges represent the number of co-mentions. - The color of the nodes represent the average sentiment of the articles they appeared in together. You can visit the website here : http://peeves.pythonanywhere.com/network
Client Project : Job and Skill Mapping Dashboard
April 1, 2023 – April 1, 2023
Skills : PostgreSQL, Relational Databases, Query Optimization, Tableau, Data Visualization The value of observing labor market trends as patterns on geographic maps has long been recognized and the ability to seamlessly zoom from state to regional and local employment patterns is taken for granted in current offerings. The aim of the project was to create an interactive map that shows job risks and possible career paths on a global scale to help communicate labor market trends in the context of specific skills and professions, and guide individuals to explore new skills, training opportunities, and jobs. Process: - Stakeholder Analysis based on insight needs - Trend analysis to visualize temporal changes - Geospatial mapping to visualize Income Parity. Databases: - ONET database : Sponsored by the U.S. Department of Labor, Employment and Training Administration - ONET Crosswalks - BLS OEWS Data : Occupational Employment and Wage Statistics by the occupational employment and Wage statistics Achievements : - Created an effective and efficient data pipeline. - Optimized SQL queries to decrease the run time from 12 to 2 Mins (-83%) using techniques like RA SQL - Created an interactive dashboard in Tableau to visualize the changes in the Job / Skill market in the US.
Object Detection | Machine Learning
October 1, 2022 – October 1, 2022
Skills : Python, Sklearn, Matplotlib, Tensorflow, Keras, NumPy, pickle The aim of the project was to recognize the object in the images, i.e., to create a multiclass classifier that decides the correct category of an unlabeled image. Dataset Description : - The training set consisted of 100 categories with labels from 0 to 99, each category containing 5000 images with a total of 500K. - The test set consisted of 100K images, which were not a part of the training set. Basic Preprocessing Steps: - Grayscale conversion - Reducing the pixel matrix size for efficient processing - Normalizing the matrix Initial KNN implementation for classification gave a base line accuracy of 44.54% After experimenting with a few other models, with varying accuracies, I decided on using a CNN model as it gave the best base line accuracy of ~60%. Advanced Pre-processing : - Data Augmentation : (rotation, zoomed, lateral shift) Final model description : - 3 Convolutional Layers with 64, 128 and 256 filters respectively. - Each followed by a Max Pooling layer and Batch Normalization layer. - Experimented with Dropouts but chose not to include it as it did not make much difference. - Neuron activation used : ReLU - Loss function : Categorical Cross Entropy (Softmax + Cross Entropy) - Callback on early stop and plateau Training epochs : 50 Batch Size : 128 Final Model Accuracy : 72.84% (2nd Overall/100)
Marketing in a Digital World
Coursera
June 25, 2026 – Present
Databases and SQL for Data Science with Python
Coursera
June 25, 2026 – Present
Digital Media and Marketing Strategies
Coursera
June 25, 2026 – Present
Digital Media and Marketing Principles
Coursera
June 25, 2026 – Present
Data Science Methodology
Coursera
June 25, 2026 – Present
Machine Learning with Python
Coursera
June 25, 2026 – Present
Marketing in an Analog World
Coursera
June 25, 2026 – Present
Python for Data Science, AI & Development
Coursera
June 25, 2026 – Present
Tools for Data Science
Coursera
June 25, 2026 – Present
Data Visualization with Python
Coursera
June 25, 2026 – Present
COVID19 Data Analysis Using Python
Coursera
June 25, 2026 – Present
Capstone: Retrieving, Processing, and Visualizing Data with Python
Coursera
June 25, 2026 – Present
Logistic Regression with NumPy and Python
Coursera
June 25, 2026 – Present
Digital Marketing Analytics in Practice
Coursera
June 25, 2026 – Present
What is Data Science?
Coursera
June 25, 2026 – Present
Digital Marketing Analytics in Theory
Coursera
June 25, 2026 – Present
Understanding Financial Markets
Coursera
June 25, 2026 – Present
Machine Learning
Stanford University
June 25, 2026 – Present
Data Analysis with Python
Coursera
June 25, 2026 – Present
Python for Data Science: Fundamentals Course
Dataquest.io
June 25, 2026 – Present
Python Basics for Data Analysis
Dataquest.io
June 25, 2026 – Present
Python for Data Science: Intermediate Course
Dataquest.io
June 25, 2026 – Present
Pandas and NumPy Fundamentals
Dataquest.io
June 25, 2026 – Present
Exploratory Data Visualization
Dataquest.io
June 25, 2026 – Present
Storytelling Through Data Visualization
Dataquest.io
June 25, 2026 – Present
Data Cleaning and Analysis
Dataquest.io
June 25, 2026 – Present
Data Cleaning in Python: Advanced
Dataquest.io
June 25, 2026 – Present
Data Cleaning Project
Dataquest.io
June 25, 2026 – Present
Spoken Language Processing in Python
DataCamp
June 25, 2026 – Present
Sentiment Analysis in Python
DataCamp
June 25, 2026 – Present
Natural Language Processing with spaCy : Advanced
DataCamp
June 25, 2026 – Present
Advanced Feature Engineering for NLP in Python
DataCamp
June 25, 2026 – Present
Large Language Models (LLMs) Concepts
DataCamp
June 25, 2026 – Present
Cultural Fit Analysis
The candidate's project diversity, ranging from data visualization and SQL optimization to advanced object detection and NLP with LLMs, indicates a broad interest and adaptability. The roles at Realbotix and IU Health align well with an ML Engineer target role, particularly with a focus on Generative AI. The academic and teaching experience also suggests a collaborative and knowledge-sharing mindset. However, the rapid progression in roles at Realbotix (Generative AI Engineer to Head of AI in less than a year) might warrant further inquiry into the scope and responsibilities of these roles to fully assess cultural fit and long-term stability.
Soft Skills & Operational Fit
The candidate's experience as an Associate Instructor and Graduate Teaching Assistant suggests strong communication and mentoring skills. Project descriptions indicate problem-solving abilities and collaboration with cross-functional teams. The progression from Generative AI Engineer to AI-Architect and Head of AI at Realbotix, albeit in a short timeframe, suggests leadership potential and rapid learning. However, without psychometric test results, a comprehensive assessment of work attitude, stress handling, and team collaboration is not possible.