sayali Moon

About

AI/ML engineer with 9 months of internship experience in machine learning, deep learning, NLP, computer vision, and generative AI. Experienced in building end-to-end AI applications using Python, LLMs, RAG, FastAPI, Streamlit, TensorFlow, PyTorch, and YOLOv8.

Experience

• AI & Data Science Intern
  Rubixe AI Solutions
  2025 – 2026

Education

• B.Tech in Computer Technology
  Kavikulguru Institute of Technology and Science, Ramtek
  Computer Technology · 2021 – 2025

Skills

• SQL
• FastAPI
• ChatGPT
• Python
• Prompt Engineering
• Data Science
• Clustering
• Classification
• Feature Engineering
• Matplotlib
• Model Evaluation
• Streamlit
• OpenCV
• Flask
• PyTorch
• TensorFlow
• Image Processing
• Seaborn
• CNN (Convolutional Neural Network)
• Keras
• LangChain
• JSON
• NLTK
• Plotly
• Spacy
• API Integration Understanding
• Workflow Automation
• Natural Language Processing
• RAG
• Multi-Agent Systems
• Object Detection
• LLMs
• Transformers
• AI Agents
• XGBoost
• AI Orchestration
• Embeddings
• Semantic Search
• Regression
• Video Processing
• Hyperparameter Tuning
• Transfer Learning
• Ragas
• Sentence Transformers
• Faiss
• Image Classification
• Gradio
• Llama
• Bert
• Recurrent Neural Networks
• Chatbots

Projects

• AI-Based Smart Exam Proctoring & Suspicious Activity Detection
  Python, OpenCV, YOLOv8, DNN/SSD, Deep Learning, Flask, Streamlit

  Developed a real-time exam monitoring system using DNN/SSD for face presence detection and YOLOv8 for multi-person and forbidden-object detection. Implemented weighted risk scoring for missing face, multiple persons, and forbidden objects to classify examination sessions based on suspicious activity. Generated automated JSON exam integrity reports with timestamped events and flagged-frame evidence, with Flask and Streamlit interfaces for session monitoring.

• AI News Event Clustering
  Python, NLP, Pandas, NumPy, Scikit-learn, Clustering, Matplotlib

  Developed an NLP-based unsupervised learning system to group related news articles into meaningful event clusters. Applied text preprocessing and feature representation to transform news content into machine-readable representations for similarity-based clustering. Analyzed generated clusters to identify related news stories and discover common event-level patterns across collections of articles.

• AI-Powered RAG System for Legal & Medical Document Q&A
  Python, LangChain, Groq LLaMA-3.1, FAISS, Sentence Transformers, RAGAS, Streamlit

  Developed an end-to-end RAG-based question-answering system using LLaMA-3.1 through Groq API, LangChain, FAISS, and Sentence Transformers for legal and medical documents. Implemented recursive document chunking with chunk size 1000 and overlap 200 for semantic retrieval and Top-K context selection. Evaluated retrieval and generation quality using RAGAS, achieving Context Precision of 0.90, Context Recall of 0.83, and Faithfulness of 0.60.

• AgentFleet – Multi-Agent AI & RAG Task Orchestration Platform
  Python, FastAPI, Streamlit, RAG, Sentence Transformers, FAISS, SQLite, LLMs

  Built a multi-agent AI system combining RAG with Research, Analysis, Worker, and Reviewer agents for document-grounded task execution. Implemented PDF, DOCX, and TXT document ingestion with chunking, embeddings, semantic retrieval, and persistent knowledge storage for document-based tasks. Developed FastAPI APIs and a Streamlit dashboard with reviewer-driven retry handling and SQLite-based tracking of task status, agent runs, attempts, duration, token usage, and estimated cost.

Courses & certifications

• AI Expert (AIE) · DataMites
• Artificial Intelligence Engineer · NASSCOM FutureSkills Prime
• Certified AI Expert · IABAC
• Certified Data Scientist · IABAC
• AI & Data Science Internship Certificate · Rubixe AI Solutions