Hi, I'm Kapil Mirchandani
A Machine Learning Engineer, passionate about building robust models and scalable serving pipelines.
About
I am a Machine Learning Engineer with two years of experience, currently studying at The University of Ottawa, Specializing in Applied Artificial Intelligence. I enjoy applying Machine Learning to end-to-end real-world tasks, including model research, as well as building inference pipelines for model deployment. I am proficient in TensorFlow and PyTorch, as well as the web frameworks Django and Flask for model serving. I have worked with a variety of Machine Learning applications, including Generative Networks, OCR,Semantic Similarity, Object Detection, Optical Character Recognition, Text Clustering and most recently, Prompt Engineering. My skills also include MLOps – this includes hosting Tensorflow models on AWS, optimizing their performance and writing the backend to call these models, and parse and store their outputs.
Education
MEng, Electrical and Computer Engineering (Concentration in Applied Artificial Intelligence)
University of Ottawa
B.E., Electronics and Telecommunication Engineering
Pune Institute of Computer Technology
Higher Secondary Education
Fergusson College
Secondary Education
The Bishop's School, Pune
Experience
- Working on the research and development of various Machine Learning algorithms used for semantic similarity, clustering, OCR and various other tasks.
- Conducting research to build custom Machine Learning models as well as evaluating pre-trained models to perform tasks associated with note taking.
- Assisting in backend system design and data platform design for deployment of these models.
- Also involved in operationalization and deployment of the researched algorithms ie. creating pipelines for training and inference of models.
- Contributed to the development of various Machine Learning solutions to help automate end user issues.
- Involved in research, scoping, benchmarking, operationalization and deployment of different NLP algorithms.
- Worked on developing Machine Learning pipelines to handle training, updation and deletion of Machine Learning models, and also inference pipelines for these models.
- Investigated and solved high priority bugs and helped improve underperforming models for clients.
- Contributed to the development of CRAL, a library used for abstraction of well known deep learning architectures for Computer Vision.
- Worked on the addition of well known deep learning architectures for computer vision into the library.
- Intensively involved in implementation, integration and testing of object detection models.
- Achieved mAP scores of more than 0.6 on standard benchmark datasets for all the integrated object detection models.
Publications
DPSRGAN: Dilation Patch Super-Resolution Generative Adversarial Networks
- Developed a novel Generative Adversarial Network architecture to increase the resolution of images.
- The input to the network is a low resolution image, which is upscaled natively by the network.
- The model is capable of upscaling input image by 4x the original resolution.
- The metrics obtained from our DPSRGAN are better than the previously proposed SRGAN, with a MOS of 3.91 out of 5 and a PSNR of 32.24.
Big Data Analytics for Sustainable Cities: Pune Tree Census Data Exploratory Analysis
- Developed a pipeline for analysis of tree census data using data of Pune, India.
- Introduced a novel metric, the Flora Biodiversity Index (FBI), to quantify the diversity of trees in a region.
- Drew insights from the data to determine uniformity of tree cover, areas deficient in trees and areas having a lower biodiversity.
- Our pipeline will be useful for cities to analyse their current green cover and work on making it better.
Projects
![music streaming app](/assets/img/project-alpr.webp)
An application that can read license plates of vehicles and register offences.
![quiz app](/assets/img/project-network_anomaly.webp)
A Machine Learning approach towards network threat detection.
![Screenshot of web app](/assets/img/project-drone.gif)
An autonomous drone for survey and rescue.
- Tools: ROS, OpenCV
- A drone capable of detecting beacons and autonomously maneuvering itself in response.
- Part of the e-Yantra Robotics Competition (eYRC) held by IIT Bombay.
Skills
Languages
![](/assets/img/python-logo-1-300x300.jpg)
![](/assets/img/cpp-logo.png)
![](/assets/img/matlab-logo.png)
![](/assets/img/java-logo.png)
![](/assets/img/Clojure-logo.png)
![](/assets/img/sql-logo.png)
Libraries
![](/assets/img/numpy-logo-1-500x500.jpg)
![](/assets/img/pandas-logo-2-500x500.jpg)
![](/assets/img/opencv-logo-1-500x500.jpg)
![](/assets/img/matplotlib-logo-1-500x500.jpg)
Frameworks
![](/assets/img/tensorflow-logo-1.png)
![](/assets/img/pytorch-logo.png)
![](/assets/img/sk-learn-logo-1-500x500.jpg)
![](/assets/img/django-logo.webp)
![](/assets/img/flask-logo.png)
Developer Tools
![](/assets/img/git.png)
![](/assets/img/docker-logo.webp)
Databases
![](/assets/img/postgresql-logo.png)
![](/assets/img/mysql-logo-1-300x300.jpg)