We have nearly always worked with numerical datasets in our standard machine learning / data science problems. Whenever we came across a column with text data, we always converted it to numeric data using data preprocessing methods such as one-hot encoding and label encoding. Let’s imagine we have a dataset with simply text data and we want to predict which twitter comments are good and which are bad; obviously, label encoding is not possible on that dataset. As a result, we use Natural Language Processing(NLP) in this case.
read moreAbout
.jpg)
AI Engineer
- City: Veldhoven, Netherlands
- LinkedIn: linkedin.com/in/adwaitkelkar
- Degree: M.Sc. Artificial Intelligence and Engineering Systems
- Email: adwaitkelkar12@gmail.com
Experienced professional with a strong foundation in state-of-the-art AI algorithms and deployment stacks, backed by hands-on industry experience. Proficient in Python and modern AI frameworks, with a focus on object detection, action recognition, and edge AI. I am currently pursuing an M.Sc. at TU Eindhoven and working on explainable AI for financial forecasting as part of my master's thesis at ASML.
Skills
Deep Learning & Computer Vision
Edge AI & Deployment
Machine Learning & Data
Programming Languages
Resume
Summary
Adwait Kelkar
Experienced computer vision professional with industry exposure in end-to-end AI pipelines, model optimization, and edge deployment. Currently pursuing M.Sc. in Artificial Intelligence and Engineering Systems at TU Eindhoven.
- Veldhoven, Netherlands
- adwaitkelkar12@gmail.com
- linkedin.com/in/adwaitkelkar
Education
M.Sc. Artificial Intelligence and Engineering Systems
September 2024 - Present
Technical University of Eindhoven, The Netherlands
B.Tech. Production Engineering
July 2019 - May 2023
Vishwakarma Institute of Technology, Pune, India
GPA: 8.71
Professional Experience
Master's Thesis Intern - ASML
December 2025 - Present
Veldhoven, The Netherlands
- Developing an explainable AI algorithm for financial forecasting models as part of master's thesis research.
AI Engineer - Tata Elxsi
Computer Vision Engineer - Tata Elxsi
June 2023 - March 2024
Bangalore, India
- Built a fully automated football highlight generation pipeline with key event detection accuracy of 80%.
- Deployed multiple object detection models on Jetson Nano using model reduction techniques, achieving 85% accuracy near real-time.
- Integrated a content moderation pipeline for movie analytics with 92% moderation accuracy.
Deep Learning Intern - Tata Elxsi
March 2022 - June 2023
Bangalore, India
- Developed an automated video labeling pipeline for action recognition with 95% accuracy and contributed to a research publication.
- Built PPE detection and tracking solution for construction safety and deployed on NVIDIA DeepStream.
Projects
Every project is an opportunity to learn, to figure out problems and challenges, to invent and reinvent.
~ David Rockwell
AI Based Advertisement
Real-time object detection pipeline for context-aware ad placement in videos, achieving 95% detection accuracy and dynamic ad triggering.
Football Highlights: AI-Driven Pipeline
Engineered an end-to-end system that converts full football matches into concise highlights, improving engagement and cutting processing time.
Project Video Demos
Hands-on outputs from selected projects.
AI Based Advertisement - Output Demo
This demo shows real-time object detection inside a live video stream. When a target object is detected, the system triggers a context-aware advertisement overlay automatically.
Football Highlights - Output Demo
This output demonstrates automated highlight generation from a full football match. The model detects key moments and compiles them into a concise highlights reel.
Trading & AI Interests
How I combine quantitative thinking, automation, and AI in my personal investment workflows.
Strategy Development
In my free time, I build and test investment and trading strategies with a strong focus on risk management, consistency, and long-term decision quality.
Trade Automation
I actively work on automating equity trading workflows so I can avoid manual trigger monitoring and execute my strategy logic in a disciplined way.
AI + Markets
I am particularly interested in merging AI with trading through signal generation, pattern detection, and explainable model-driven decision support.
Outside Work
I recharge by being active and spending time in nature.
🚴
Cycling
I enjoy cycling through nature trails and open roads.
🌇
Sunsets & Nature
Being around nature and watching sunsets helps me reset and stay calm.
🏸
Squash
I play squash regularly for agility, focus, and competitive fun.
🏃
Running
Running keeps me disciplined and improves stamina and mental clarity.
TECHNICAL BLOGS
Hypothesis testing is a statistical approach that assists researchers in determining the validity of their theories. It is frequently used in statistics and data science to determine whether an event has occurred or will occur based on previous happenings.
read more
Let’s understand this concept with a simple example. You want to apply for US University to pursue Master’s degree. Then the factors on which whether you will get an admit letter from a particular University depends on the following factors i.e. GRE Score, TOEFL Score, Number of research papers published, SOP and Letter of recommendation so, basically speaking in data science terms these are your independent variables and the chances of you being admitted into an university is your dependent variable which you predict based on your independent variables. In simple terms the thing which you need to predict falls under dependent variable and the factors on which your prediction is based on are your independent variables.
read more