Shubham Kanojia
ML EngineerAI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
CTO at RoadVision AI | Agentic AI + Digital Twin Infrastructure Systems | GenAI, CV, MLOps | Driving Autonomous Intelligence at Scale
At RoadVision AI, strategic AI initiatives come to life under my stewardship as CTO, where I harness Azure's suite of AI tools to spearhead innovation. My recent transition from an AI Solutions Architect role at Indika AI to CTO at RoadVision AI signifies a commitment to leading-edge technology and impactful leadership. My background from IIT Palakkad, with a focus on Data Science, fuels my approach to data-centric solutions and machine learning architectures. My core competencies lie in crafting intelligent systems using Azure Document Intelligence and GPT Vision, coupled with Cosmos DB for robust data management. These technical proficiencies have been pivotal in transforming complex data into actionable insights, thus driving RoadVision AI's strategic direction. Our team's collaborative efforts consistently yield sophisticated AI solutions that resonate across industry verticals.
Indian Institute of Technology, Palakkad
Master of Technology - MTech, Data Science
August 1, 2021 – July 1, 2023
RoadVision AI
CTO
December 1, 2023 – Present
Delhi, India · On-site
Indika AI
AI Solutions Architect
June 1, 2023 – December 1, 2023
New Delhi, Delhi, India · On-site
Atoll Solutions
Data Science Intern
May 1, 2022 – August 1, 2022
Indian Institute of Technology, Palakkad
Placement Coordinator
March 1, 2022 – May 1, 2023
Indian Institute of Technology, Palakkad
Teaching Assistant
June 1, 2021 – May 1, 2023
ConveyInspect
Founder
July 1, 2018 – September 1, 2021
Lucknow, Uttar Pradesh, India
Automatic Visual Inspection of biscuit packaging using EDGE devices
September 1, 2022 – May 1, 2023
Project Overview: This project developed a prototype model for automatic biscuit packaging inspection. The system was designed using an EDGE device, USB cameras, and a small DC source-controlled conveyor belt, and utilizes the YOLO V5 model for defect detection. Project Goals: The goals of this project were to: Develop a prototype model for automatic biscuit packaging inspection Evaluate the performance of the prototype model Identify potential areas for improvement Project Methods: The project team used the following methods to develop the prototype model: Hardware selection: The project team selected an EDGE device, USB cameras, and a small DC source-controlled conveyor belt. Software development: The project team developed a software application that uses the YOLO V5 model to detect defects in biscuit packaging. Data collection: The project team collected a dataset of biscuit packaging images with and without defects. Model training: The project team trained the YOLO V5 model on the collected dataset. Model evaluation: The project team evaluated the performance of the trained model on a held-out test dataset. Project Results: The project team successfully developed a prototype model for automatic biscuit packaging. Please follow this link for validation output: Please follow this link for validation output: https://drive.google.com/drive/folders/1vfe6g-KdbiudqWsXpAVt8rcbJqMOHgkh?usp=sharing
Multiple Object Tracking On Videos
August 1, 2021 – November 1, 2021
Project Goal : The goal of this project is to develop an object tracking algorithm using the DeepSORT algorithm. The DeepSORT algorithm is a learning-based algorithm that can track multiple objects in real-time. Project Approach: The DeepSORT algorithm will be trained on a dataset of car images and videos. The training data will be used to learn the appearance and motion features of cars. The trained DeepSORT algorithm will then be used to track cars in real-time. Project Outcomes: The successful development of the DeepSORT algorithm for car tracking will have a number of potential benefits. These benefits include: Improved traffic safety: The ability to track cars in real-time can be used to improve traffic safety by detecting and tracking potential hazards, such as vehicles that are driving erratically or that are about to collide with other vehicles. Improved traffic efficiency: The ability to track cars in real-time can be used to improve traffic efficiency by tracking vehicles and predicting their future movements. This information can be used to optimize traffic signals and to provide real-time traffic updates to drivers. Improved law enforcement: The ability to track cars in real-time can be used to improve law enforcement by tracking vehicles that are suspected of being involved in criminal activity. Project Mitigation Strategies The following mitigation strategies will be used to address the risks associated with this project: The data collection and preparation process will be carefully planned and executed. The DeepSORT algorithm will be trained on a large and diverse dataset of car images and videos.
Amazon Reviews Classification System using LSTM Model
August 1, 2021 – September 1, 2021
I am excited to share my latest project, where I tackled the challenge of analyzing and processing a vast dataset of customer ratings on Amazon reviews spanning a period of 18 years. With approximately 35 million reviews up until March 2013, the goal was to develop a Python-based solution that could effectively preprocess the data, train models, and make accurate predictions. In this project, I delved into various probabilistic models such as Naïve Bayes and Multinomial Naïve Bayes. Additionally, I explored decision models like Random Forest and implemented a deep learning model using LSTM (Long Short-Term Memory) to achieve an impressive test data accuracy score of 92%. To enhance the usability and accessibility of the project, I built a user interface utilizing Dash. This interface proved to be instrumental in enabling seamless data analysis, exploration, visualization, modeling, instrument control, and reporting. By undertaking this project, I not only gained extensive hands-on experience in Python for data preprocessing and model training but also sharpened my skills in implementing and fine-tuning various machine learning algorithms. The combination of robust model performance and a user-friendly interface makes this project a valuable asset for businesses seeking to analyze customer ratings effectively. I am thrilled to share more details about my project and discuss how it can contribute to enhancing data analysis and decision-making processes. Feel free to reach out to me for further information or collaboration opportunities.
Data Science & Machine Learning Complete Course
Coding Ninjas
June 23, 2026 – Present
Introduction to Machine Learning
Indian Institute of Technology, Palakkad
June 23, 2026 – Present
Data Structures and Algorithm in JAVA
Coding Ninjas
June 23, 2026 – Present
Introduction to Programming Using Java
Coding Ninjas
June 23, 2026 – Present
Cultural Fit Analysis
The candidate's diverse project portfolio, ranging from computer vision to NLP, demonstrates adaptability and a broad interest in ML applications. The entrepreneurial background suggests a proactive and innovative mindset, which could be a strong cultural fit for dynamic environments. However, the lack of explicit team-based project descriptions makes it difficult to fully assess collaboration style. The target role of ML Engineer aligns well with the candidate's technical background and project experience.
Soft Skills & Operational Fit
The candidate's experience as a Founder and CTO, along with a Placement Coordinator role, suggests strong leadership, communication, and organizational skills. The project descriptions indicate a structured approach to problem-solving and an understanding of project goals and mitigation strategies. These traits are beneficial for operational fit in a senior ML Engineer role.