Overview
As an AI/ML Engineer Lead at Metafiser, under the guidance of Dr. Vaseem Durani, I developed AI-powered solutions and gained hands-on experience in machine learning, recommendation systems, and client-oriented AI services.
I collaborated with the team to build a Spotify Music Recommendation System using Python, Google Colab, and the Spotify Web API.
My work involved data preprocessing, feature engineering, model training, and implementing Content-Based and Collaborative Filtering techniques, with the vision of combining both into a Hybrid Recommendation System.
This experience strengthened my technical expertise while providing valuable exposure to real-world AI development, research, teamwork, and delivering AI solutions aligned with business needs.
Impact
1. AI Solution Development & Client Requirements
Problem
Clients required AI-driven solutions tailored to their business needs with clear technical feasibility.
Solution
Analyzed client requirements, explored AI use cases, and designed machine learning solutions aligned with business objectives.
Impact
- Improved understanding of real-world AI applications.
- Delivered solutions aligned with client expectations.
- Enhanced ability to translate business problems into AI solutions.
2. Spotify Music Recommendation System
Problem
Users struggled to discover personalized music matching their listening preferences.
Developing Content-Based and Collaborative Filtering required integrating recommendation techniques with Spotify API data.
Solution
Developed a hybrid Spotify Music Recommendation System using the Spotify API, machine learning algorithms, and recommendation techniques.
Impact
- Improved recommendation accuracy through personalized suggestions.
- Successfully integrated real-time music metadata using Spotify APIs.
- Strengthened expertise in recommendation systems and API integration.
Personal Learning
Working on the Spotify Music Recommendation System gave me practical experience in recommendation engines, API integration, and machine learning workflows.
I learned how Content-Based and Collaborative Filtering complement each other and explored combining them into a Hybrid Recommendation System to deliver highly personalized recommendations based on a user's listening history, preferences, and music taste.
3. Machine Learning Model Development
Problem
Building accurate recommendation models required effective data processing and feature engineering.
Solution
Performed data preprocessing, feature extraction, model training, and evaluation using machine learning techniques and finally deployed the solution on Streamlit.
Impact
- Improved model performance with optimized preprocessing.
- Developed practical experience in the end-to-end ML pipeline.
- Built scalable and reusable machine learning workflows.
4. AI Service Delivery & Technical Consulting
Problem
Many organizations lacked awareness of how AI could solve operational challenges.
Solution
Provided AI-driven solutions, conducted workshops, and offered technical guidance to help organizations leverage AI for their specific needs.
Participated in discussions on AI service offerings, solution demonstrations, and communicating technical concepts to potential clients.
Impact
- Enhanced organizational understanding of AI applications and benefits.
- Improved technical communication and presentation skills.
- Facilitated the adoption of AI technologies across various business units.
- Learned how AI products are positioned for businesses.
- Bridged the gap between technical development and client requirements.
5. Team Collaboration & Research
Problem
AI projects required continuous experimentation, collaboration, and rapid iteration.
Solution
Worked closely with mentors and teammates under the guidance of Dr. Vaseem Durani, conducting research, testing models, and improving recommendation accuracy.
Impact
- Enhanced teamwork in AI product development.
- Strengthened research and analytical thinking.
- Gained hands-on experience in industry-standard AI development practices.