Building production-ready AI systems that solve real-world healthcare problems.
AI Engineer at MTBC CareCloud Inc. with 1+ years of experience building production-ready AI systems, data pipelines, and ML solutions focused on healthcare. I enjoy turning complex problems into reliable, scalable applications.
- 🏢 Currently: AI Engineer at MTBC CareCloud Inc.
- 🎓 Education: BS Computer Engineer — National University of Science and Technology (NUST), Islamabad
- 🌱 Focus Areas: Agentic AI, Machine Learning, Deep Learning, MLOps, Computer Vision, NLP
- 💬 Ask Me About: Healthcare AI, MLOps, BERT/NLP, Django, Python, Model Deployment
- 📫 Reach Me: 2020n02900@gmail.com
- ⚡ Fun Fact: I debug best at 2 AM with chai ☕
- 🤖 Building agentic AI workflows for healthcare at CareCloud
- 📊 Exploring LLM evaluation frameworks (BERTScore, RAGAS)
- 🚀 Learning MLOps with cloud platforms (AWS SageMaker)
- 🧠 Researching transformer architectures for medical NLP
Healthcare-focused AI · Model Deployment · Data Pipelines · Applied ML
- Building production ML systems for healthcare workflows
- Designing and deploying scalable data pipelines
- Collaborating with cross-functional teams on AI-driven product features
Backend APIs · Data-Driven Applications · Cloud Technologies
- Built backend services and RESTful APIs using Django and FastAPI
- Developed data-driven applications with modern web frameworks
- Worked across the full development lifecycle from design to deployment
Python · Whisper · Plugin Architecture · Speech Recognition · Agentic AI
Modular, plugin-based voice assistant powered by OpenAI Whisper for speech recognition, rule-based intent classification, and offline text-to-speech. Features a Tkinter GUI and extensible plugin system — demonstrating agentic AI patterns and production-ready architecture.
PyTorch · Computer Vision · Weak Supervision · Transfer Learning · Semantic Segmentation
Research-grade deep learning project training a semantic segmentation model on the LoveDA land-cover dataset using only sparse point annotations instead of full pixel masks. Implements a custom Partial Cross-Entropy (pCE) loss — addressing real-world annotation workflows where labeling every pixel is impractical.
Python · LLM · Retrieval-Augmented Generation · Vector Search · NLP
Hands-on exploration of multiple Retrieval-Augmented Generation architectures — comparing different retrieval strategies, embedding approaches, and orchestration patterns. Practical reference for building modern LLM applications with grounded, reliable responses.
Python · Multimodal AI · Agent · NLP · Hackathon Project
AI-powered system built for the HackerRank Orchestrate 24-hour hackathon that decides which WhatsApp messages deserve immediate attention, which should wait, and which to mute. Reasons over multimodal inputs — text, images, and voice notes — to triage noisy message streams intelligently.
Python · BERT · NLP · Healthcare AI · Evaluation Metrics
Evaluated conversational AI responses in medical chatbots using BERTScore semantic similarity metrics and model analysis. Built an automated evaluation pipeline to ensure clinical response quality and relevance for healthcare AI applications.
Django · XGBoost · Machine Learning · Web Application · Full-Stack ML
Django-based web application that uses XGBoost to predict optimal tips from dining context — party size, time of day, cuisine, and more. Full-stack ML product covering model training, API design, and deployment.
National University of Science and Technology (NUST), Islamabad
🎓 BS Computer Engineer
I'm always open to discussing interesting projects, especially in:
- 🏥 Healthcare AI — Clinical NLP, medical imaging, patient data systems
- 🤖 Agentic AI — Autonomous workflows, LLM agents, tool use
- 📚 Machine Learning — Model deployment, MLOps, evaluation
- 🧠 Deep Learning — Transformers, computer vision, NLP research
Open to: Full-time opportunities, freelance projects, research collaborations, and open-source contributions.
💡 "Turning complex problems into reliable, production-ready AI systems."
⭐️ From MeerBaloch7