Shahid Afridi
Highly skilled in machine learning, data analysis, and cyber security, with a goal to contribute to innovative projects and enhance the efficiency of processes through the use of advanced technology.
- Role
ML & Python Developer
- Years of Experience
3.3 years
Skillsets
- Deepseek
- pandas
- Scikit-learn
- Sentence transformers
- SQL
- Tableau
- TensorFlow
- Tkinter
- Word embeddings
- Myscale
- Celery
- Claude
- NumPy
- Excel
- Gemini
- GPT
- Grok
- LLAMA
- PubSub
- PyQt5
- Rags
- Time Series Analysis
- vector embeddings
- Python - 3.0 Years
- NLP
- Computer Vision
- LLMs - 3.0 Years
- OpenCV
- EDA
- voice cloning
- Bert transformers
- Data Analysis
- Vector databases
- voice cloning
- LangChain - 3.0 Years
- Data Analysis
- BeautifulSoup
- C
- ChromaDB
- Flask
- Jupyter Notebook
- Keras
- Machine Learning
- Matplotlib
- Mediapipe
- NLTK
Professional Summary
3.3Years
- Apr, 2023 - Jul, 20252 yr 3 months
Machine Learning Engineer
GeeksForGeeks - Jan, 2022 - Apr, 2022 3 months
Associate Analyst
Goldstone Technologies - Jun, 2021 - Jul, 2021 1 month
Python developer Intern
Celebal Technologies - Dec, 2020 - Jan, 2021 1 month
Trainee
Oil and Natural Gas Corporation - May, 2021 - Jun, 2021 1 month
Data Science and Business Analytics Intern
The Sparks Foundation
Applications & Tools Known
.png)
Flask
Tableau
Tableau Prep
Jupyter Notebook
Excel
Tableau
Jupyter Notebook
PyQt5
Excel
Work History
3.3Years
Machine Learning Engineer
GeeksForGeeksApr, 2023 - Jul, 20252 yr 3 months
Setup the entire ML Backend on Flask while creating RESTful APIs. Engineered a RAG global search system on the GeeksForGeeks platform, facilitating searches on products through textual queries, while replacing Google's ElasticSearch and reducing costs. Introduced a premium AI Chatbot feature enabling dynamic question answering on any visited articles page, improving articles understanding with real-time doubt clearance using AI. Pioneered a system using AI technologies to transform textual scripts into short lecture videos, featuring a human avatar for narration and text highlighting. Streamlined article review processes by automating grammar checks, formatting assessments, and error identification, thus eliminating manual intervention and increasing efficiency. Engineered a feature to generate problem statements, examples, and editorial content for coding problems, employing Language Models, thus eliminating manual writing of the content and increasing efficiency. Developed a coding problem Recommendation System for DSA articles, streamlining the process by eliminating manual intervention in mapping relevant problems to articles. Integrated a premium feature for generating concise and specific summaries of articles, enhancing readability and user experience on the platform. Developed an automated proctoring system for contests to effectively detect and prevent cheating by participants. Integrated a queuing mechanism for chatbot API to improve the throughput using Google Pub/Sub and Celery.
Associate Analyst
Goldstone TechnologiesJan, 2022 - Apr, 2022 3 months
Building visualizations, dashboards using Tableau. Data modelling with Tableau Prep. Data analysis using SQL and Tableau.
Python developer Intern
Celebal TechnologiesJun, 2021 - Jul, 2021 1 month
Developing a blood bank management system using python, SQL, PyQT5.
Data Science and Business Analytics Intern
The Sparks FoundationMay, 2021 - Jun, 2021 1 month
Building prediction models using supervised and unsupervised learning.
Trainee
Oil and Natural Gas CorporationDec, 2020 - Jan, 2021 1 month
Learning NLP while analysing oil drilling datasets from ONGC.
Achievements
- Setup the entire ML Backend on Flask
- Engineered global search system on the GeeksForGeeks platform
- Introduced premium AI Chatbot
- Pioneered system transforming scripts into lecture videos
- Streamlined article review processes
- Developed coding problem Recommendation System
- Integrated premium feature for article summaries
- Social work volunteer at Nasscom Foundation
- Sports Club member and football team member at UPES
Major Projects
12Projects
Global Search System
A searching feature to get the most relevant object from GeeksForGeeks products such as Articles, Coding problems, Jobs, Courses. Utilised LangChain for similarity search with retrieval chains and generated embeddings using ChromaDB on sentence transformers. Created a Flask API for integration on GeeksForGeeks webpages.
Interactive AI Chatbot for articles
LLM based Chatbot for question answering on any articles page. Generated embeddings of articles using sentence transformers and performed vector searches in a vectorDB. Used prompt engineering for custom prompts and LLM interaction.
Contest proctor
An automated proctoring system to detect malpractices by contestants. Utilised Yolo v7 for object detection and LLMs to identify potential malpractice from video streams.
Script to Video Generation
Generated videos featuring a human avatar narrating input text with text highlighting. Used Google Text-to-Speech, ElevenLabs API for voice cloning, Whisper for timestamps, Wav2Lip for lip syncing, and moviepy for text highlighting.
Content Review
Implemented an automated article content review system with predefined checks. Used ChromaDB and sentence transformers for searching similar titles, Gramformer for grammar checking, and NLTK/textstat for difficulty analysis.
Problem Statement and Editorial Generation
Generated problem statements, examples, and editorial content for coding problems using GPT-3.5 and prompt engineering for HTML formatting.
DSA Problems Recommender for Articles
Developed a recommendation system to suggest problems for DSA articles using embeddings and cosine similarity between articles and problems.
Summary Generator
Crafted concise summaries of articles using OpenAI and Google's LLMs with prompt engineering and LangChain for custom formatting.
AI Gym Trainer
An automated workout exercise analyzer using mediapipe to compute angles and assess exercise precision.
AI Gesture Volume Controller
Developed a device volume controller utilizing hand gestures, powered by mediapipe, OpenCV, and Numpy for gesture recognition and volume adjustment.
Financial Textual Classification
Fine-tuned BERT-Transformers for classification of financial textual data from financial reports.
Face detection and face recognition
Implemented face detection and recognition systems by fine-tuning YOLO and VGGNet using transfer learning for real-time detection from live feeds.
Education
B. Tech in CSE - Cyber Security and Forensics
University of Petroleum and Energy Studies (2022)Senior Secondary (CBSE)
Montfort School, Roorkee (2017)Secondary (CBSE)
Montfort School, Roorkee (2015)Google IT automation with python
CourseraPython with Machine Learning
Ducat, Noida (2021)
Certifications
Python with machine learning
Google it automation with python