image

Muhammad Azam Afridi

Machine Learning Engineer
azamafridi.ai@gmail.com   +92 (325) 9594992

About Me

I am currently working as a Machine Learning Engineer at 99 Technologies Pvt. Ltd. Prior to this, I was an undergraduate student at the Ghulam Ishaq Khan Institute of Engineering Sciences and Technology (GIKI), where I graduated with a Bachelor's degree in Artificial Intelligence. I was awarded a gold medal for being the highest-ranked student in my faculty. Additionally, I served as a research assistant at the Artificial Intelligence (AI) Lab under the supervision of Dr. Khurram Khan Jadoon. My final year project involved developing a multilingual video translator and a video-based question-answering system using multimodal AI models.

I have a keen interest in natural language processing and data science, and I am particularly passionate about applying cutting-edge AI techniques to tackle the evolving challenges of real-world applications. Through my academic achievements and professional experience, I am committed to fostering innovation and enhancing the practical impact of AI solutions.

Experience

May 2024 - Present
99 Technologies Pvt. Ltd

Machine Learning Engineer

  • Optimized Selenium-based data collection pipeline, transforming it into a scalable, parallel processing system, reducing web scraping time by 18%.
  • Developed and deployed a custom NLP chatbot system using Retrieval-Augmented Generation (RAG) technique, enabling accurate question answering related to tech products on the company's website.
  • Developed and integrated a personalized recommendations system utilizing semantic search, providing customized recommendations based on user queries and behavioral patterns.
April 2023 - May 2024
Artificial Intelligence (AI) Lab

Research Assistant

  • Conducted research on developing scalable and accurate systems for multilingual video translation and video-based question answering at the Artificial Intelligence (AI) Lab under the supervision of Dr. Khurram Khan Jadoon.
  • Fine-tuned the Helsinki-NLP/opus-mt-en-ar model using the Web Inventory of Transcribed and Translated Talks (WIT3) dataset, improving the BLEU score by 6 points for more accurate translation of scientific terminology.
  • Integrated OpenAI's GPT-4 to leverage its multimodal capabilities, enabling advanced visual question answering and enhancing the system's ability to process and respond to complex queries involving both text and images.
Jun 2023 - Aug 2023
Kavtech Solutions (Private) Ltd

Python Intern

  • Conducted research on implementing Convolutional Long Short-Term Memory (Conv LSTM) networks for Human Action Recognition, enabling precise action identification in video sequences.
  • Acquired and curated a diverse dataset of human actions using roboflow in video format to support model training.
  • Trained the model on the dataset, optimizing accuracy in action classification and achieving 77% accuracy on the test set.
Jan 2021 – Jun 2021
GIK Institute of Engineering Sciences and Technology (GIKI)

Teacher Assistant

  • Assisted the professor in preparing and delivering course material for CE121 (Digital Logic and Design) to a class of over 65 students.
  • Graded assignments and exams, providing detailed feedback to improve student understanding and performance.
  • Conducted review sessions before exams to help students consolidate their knowledge and address questions.

Projects & Publications

Projects

  • Multilingual Video Translator and Video Based QA System GitHub

    Designed and implemented a multilingual video translation & Chatbot system using advanced AI models, enabling users to translate videos from non-native languages into their native language. Additionally, integrated a context-aware chatbot that responds to user queries based on information extracted from the uploaded videos.

    • Integrated OpenAI’s Whisper for speech-to-text, Meta’s NLLB model for text translation, and Microsoft Azure’s Speech API for multilingual speech translation.
    • Fine-tuned the T5 model to enhance text translation from English to French, resulting in a 17% improvement in BLEU score.
    • Developed an intelligent chatbot using OpenAI GPT 3.5 LLM, employing RAG (Retrieval augmented generation) technique for context-driven answers.
    • Deployed the application on Azure’s nct4-v3 Virtual Machine.

  • Custom AI-Powered Face Recognition System Using FaceNet and MTCNN GitHub

    Developed a custom face recognition system using cutting-edge AI models, achieving over 86% accuracy in facial recognition tasks. The project utilized FaceNet and MTCNN to build a robust and reliable solution for real-time facial detection and recognition.

  • Driver drowsiness detection application using CNN & YOLO GitHub

    Developed an AI-driven driver drowsiness detection system utilizing Convolutional Neural Networks (CNNs) with ResNet architecture and the YOLO v5 algorithm to assess drowsiness based on facial features, achieving 73% and 81% accuracy, respectively.

Publications

  • Enhancing Technical Terminology Accuracy in English-to-Arabic Speech Translation
    Muhammad Azam Afridi, Rajab Ali, Khurram Khan Jadoon
    Submitted to IEEE DataSciMI  

Awards & Accolades

  • Jun 2024

    Faculty Gold Medal (Jun 2024)

    Received the Faculty Gold Medal for best academic performance among all candidates for the degree of Bachelor of Science in Artificial Intelligence at GIK Institute.

  • 2020 - 2024

    Dean's Honor Certificates

    Received the Dean's Honor Certificate, awarded to students with a distinction in an academic semester, in six out of eight semesters.

  • May 2024

    Lab2Market Project Competition Winner

    My project, NatLearn, secured 1st position in the Lab2Market competition at the Artificial Intelligence Expo and Conference for the best AI-based project at Pak-Austria Fachhochschule Institute of Applied Sciences and Technology, winning a prize of 100,000 PKR. (approximately 500 USD).

  • March 2024

    Numaish Competition Winner

    My final year project, Multilingual Video Translator and RAG Chatbot, won the Numaish Competition sponsored by Motive for the Best Computer Science Domain Project Award across Pakistan at NaSCon, FAST NUCES Islamabad, Pakistan (2024), earning a cash prize of 50,000 PKR. (approximately 250 USD).

  • Sep 2022

    GIKI Alumni Association FA Merit (Scholarship)

    Received the GIKI Alumni Merit Scholarship worth 400,000 PKR (approximately 1,500 USD) for the 2022-2023 session, awarded for excellent academic performance.

Education

  • BS in Artificial Intelligence
    GIK Institute of Engineering Sciences and Technology (GIKI), Pakistan
    Sep. 2020 - June 2024
    CGPA: 3.75/4.00 (Gold Medalist)
    Thesis: Multilingual Video Translator & Chatbot System using Large Language Models (LLMs)
  • Higher Secondary School Examination
    Pakistan School and College Salalah, Oman
    Mar. 2018 - Apr. 2020
    Final grade: A1

Research Interests

  • Natural Language Processing (NLP)
  • Large Language Models (LLMs)
  • Computer Vision
  • Machine Learning
  • Deep Learning

Technical Skills

  • Python
  • Tensorflow
  • Huggingface
  • Langchain
  • Opencv
  • Django
  • Selenium

Languages

  • English: IELTS Academic Band 8
  • Urdu (Native)