Data R&D • Cyclist • Dev.py
[m] Zaid
[a] Learner more obsessed with "what if" than "how to" — ideas first, code (eventually).>_ aboutMe_
This is my digital footprint — Welcome!
I'm a curious soul who loves figuring things out and diving deep into stuff that makes me go "whoa!" – especially when it comes to AI, soft dev, and data science.
I’ve been building and breaking things in code ever since I graduated in 2022, having clocked over 4 years of hands-on dev experience whether it’s a fun side project or a tricky problem, i’m always down to explore, experiment, and share what i discovered.
When i’m not glued to my screen, you’ll probably find me trekking through trails, diving into research, or experimenting with ideas with my assistant Mr GPT.
>_ myProjects_
Lookalike - Celebrity Face Matching ↗
Built a web-based application that matches user-uploaded photos to celebrity lookalikes using facial recognition. Utilized FastAPI, DeepFace (ArcFace model), and FAISS to search over 26K+ celebrities with real-time matching. Implemented cosine similarity for precise score-based matching, delivering high-accuracy results and an interactive user experience.
Automated Social Media System
Developed a fully automated pipeline that posts AI-generated images daily across Telegram ↗, Instagram ↗, and X (Twitter) ↗. Scraped high-quality images from Lexica.art using Playwright, and captioned posts with prompts via Python automation scripts. Integrated platform APIs and scheduling to ensure fully hands-free operation and consistent multi-channel growth.
Air Quality and Pollution Assessment ↗
Developed a machine learning model to predict air quality levels using environmental and demographic data. Achieved 90% accuracy with a Random Forest classifier after analyzing 5,000+ samples and performing extensive feature engineering. Evaluated additional models including Gradient Boosting (GBM) and Support Vector Machines (SVM) to support environmental monitoring and public health planning.
2029 Lok Sabha Election Prediction ↗
Built a machine learning pipeline to predict popular vote % and seats won for major Indian parties like BJP, INC, and AAP in the upcoming elections. Using historical election data, trained Random Forest and XGBoost models with achieving an average error of ±2.1% in vote share and ±4 seats.
>_ myExperience_
Business & Revenue Analyst – Smartivity Labs
New Delhi · Contractual · March - May 2025
Conducted in-depth market research for STEM toys, analyzed
packaging performance metrics (CTR, conversion, returns), and
proposed design optimizations.
Mapped concept selection processes and suggested a standardized research framework.
AI Python R&D – Cynapto Technologies
Remote · Freelance · April – May 2025
Built and deployed Flask-based REST APIs for image uploads
and facial recognition.
Integrated a PyTorch inference pipeline for real-time face detection and feature
extraction.
Associate Software Engineer – Techademy (IIHT)
Bengaluru · Full-time · June 2023 – April 2024
Enhanced the Techademy LMS platform by optimizing key features and improving user experience
using Python.
Worked collaboratively across teams to deliver scalable improvements.