Welcome to my GitHub profile! I'm Hasna Akbar Ali, an AI/ML Engineer specializing in building production-ready AI systems using LLMs, RAG architectures, FastAPI, and scalable data pipelines.
I enjoy turning complex, real-world problems into intelligent, deployable solutions — from document intelligence and AI agents to data-driven automation platforms.
- 🤖 AI/ML Engineer with hands-on experience in LLMs, RAG systems, and AI automation
- 🧠 Strong background in Machine Learning, NLP, Deep Learning, and Time Series Forecasting
- ⚙️ Experienced in building FastAPI backends, async pipelines, and NoSQL architectures
- 🌍 Worked on real-world datasets including large-scale scraping, document processing, and customer data
- 🇦🇪 Actively exploring AI Engineer opportunities in Dubai
- 💼 Open to AI/ML collaborations, research-driven projects, and impactful product development
- Python (NumPy, Pandas, Scikit-learn, TensorFlow)
- FastAPI, REST APIs, Async Processing
- Dart, Flutter (Mobile Development)
- Supervised & Unsupervised Learning
- Feature Engineering & Model Evaluation
- Deep Learning (CNNs, LSTM)
- Time Series: ARIMA, SARIMAX, Prophet, LSTM
- RAG (Retrieval-Augmented Generation)
- LLM Orchestration (GPT-4, Gemini, Groq / Llama 3)
- Prompt Engineering & Optimization
- Text Classification, Summarization, Semantic Search
- MongoDB, SQL Server
- Milvus Vector Database
- Git, GitHub, Docker, CI/CD
- Web Scraping: Selenium, Playwright
- Firebase, GraphQL
- Built an AI-powered document processor using Python & FastAPI
- Automated structured data extraction from PDFs using LLMs and Vision models
- Designed a RAG-ready pipeline with async job tracking and NoSQL outputs
- Developed an intelligent HR assistant using FastAPI and local LLMs (Ollama)
- Enabled semantic search over employee skills using SentenceTransformers
- Built a carousel-style frontend to display ranked candidate results
- Performed time series analysis using ARIMA, SARIMAX, Prophet, and LSTM
- Achieved superior performance with LSTM
- MAE: 1.60
- RMSE: 1.89
- Addressed class imbalance via augmentation and class weighting
- Implemented VGG16 and ResNet50v2
- Achieved 66% overall accuracy across 7 emotion classes
- Built an AI agent that automated ~80% of customer queries
- Integrated decision logic, ticket routing, and backend workflows
B.E. in Computer Science
P.A. College of Engineering, Mangalore
CGPA: 8.55 / 10
- IBM Certified Data Science & AI (2024)
- 🔗 LinkedIn: Hasna Akbar Ali
- 📧 Email: [email protected]
Feel free to explore my repositories or reach out —
I’m always excited to collaborate and build intelligent systems 🚀