INTERNSHIPS

Description:
At Systems Limited, I build enterprise-grade AI tools across computer vision, NLP, and generative models that enhance workflow efficiency, ensure compliance, and deliver real-world impact.

Key Responsibilities:

• Medical Claim App:
  Built an Azure OCR → OpenAI pipeline to extract handwritten prescription data, interpret it semantically, and validate each claim against internal health policy rules.
  
  

• Chat-Based Shopping Assistant:
  Developed a full-stack intelligent e-commerce platform inspired by Khaadi using React, FastAPI and Azure OpenAI. Integrated multi-agentic LLMs for natural language interaction and autonomous task flow. Currently enhancing it into a context-aware recommender system combining structured product data with user-generated content. Leveraged LangChain, vector-based RAG and user behavior modeling to simulate human-like engagement and make dynamic recommendations across browsing, cart and checkout experiences.
  
  

• Fake News Detection System:
  Created a text-classification machine-learning pipeline (Naive Bayes, SVM, Random Forest, etc.) with TF-IDF preprocessing, deployed via Streamlit frontend and Flask /predict API.
  
  

• Visual Defect Detection System:
  Fine-tuned a PyTorch ResNet50 model on the MVTec AD dataset and delivered a Streamlit app for live defect detection with performance metrics.
  
  

• Executed data preprocessing, exploratory analysis, and feature engineering with Python, Pandas, scikit-learn, TensorFlow, Keras, and PyTorch.
  
  

• Utilized Azure Machine Learning, Cognitive Services, and OpenAI to deploy models securely and monitor lifecycle health.
  
  

• Collaborated in Agile sprint cycles with cross-functional teams to deliver ethical, compliant, and scalable AI systems.
  
  

Skills Gained:

• Expertise in full-stack AI pipeline development for OCR, NLP, CV, and generative AI applications.
  
  

• Deep technical proficiency in Python, Pandas, scikit-learn, TensorFlow, Keras, and PyTorch.
  
  

• Strong experience in enterprise ML deployment & MLOps using Azure platforms.
  
  

• Ability to design and build interactive UIs and APIs using React, Streamlit, and Flask.
  
  

• Skilled in model evaluation and explainability techniques (accuracy, F1, confusion matrix, Grad-CAM, SHAP).
  
  

• Agile collaboration skills ensuring solutions meet business and compliance needs.
  
  

Project Link:

AI-Powered Medical Claim App (Click to view the project) 

AI‑Powered Chat‑Based Shopping Assistant  (Click to view the project) 

Fake News Detection System (Click to view the project) 

Visual Defect Detection System (Click to view the project) 

Project Description:

Developed an augmented reality (AR) application that simulates real-world electronic circuits, integrating AI for detection and overlaying virtual circuit models onto physical objects. The goal of the project was to provide a more interactive and engaging method for students to learn about electronics and circuit components.

Key Responsibilities:

• Implemented AR functionality using Unity to superimpose virtual circuit components on real-world objects through a mobile app.

• Integrated AI-based detection algorithms to automatically recognize electronic components and trigger corresponding AR visualizations.

• Worked on the development of the user interface (UI) to allow seamless interaction between users and the AR circuit simulation.

• Focused on creating accurate, real-time AR simulations that enabled users to interact with virtual representations of electrical circuits overlaid on physical circuits.
  
  

Skills Gained:

• Expertise in AR development with Unity and C# for interactive mobile applications.

• Hands-on experience in AI integration for object detection in AR applications.

• Enhanced problem-solving and debugging skills within Unity and AR-related technologies.

Project Link:

AR Circuit Simulator with AI (Click to view the project)