I’m a technology enthusiast with a focus on Artificial Intelligence (AI), DevOps, and Data Science. With a strong drive for innovation, I’m passionate about building impactful, real-world solutions using cutting-edge technologies. I’m constantly seeking to expand my skill set and collaborate on projects that push the boundaries of what’s possible in the digital world.
I am an active researcher in the field of AI, with a focus on Human Activity Recognition (HAR), Agriculture 4.0, and Generative Models using GANs and LLMs.
🔗 See all my publications on Google Scholar
-
An Approach for Data Augmentation in HAR with Wearable Sensors Using TIMEGAN ICML 2024 (Miami, USA) — Focuses on generating realistic synthetic data from wearable sensors using TimeGAN.
-
Using Mobile Edge AI to Detect and Map Diseases in Citrus Orchards International Journal — A systematic study of DL architectures applied to disease detection and fruit counting in orchards.
- Prompt Engineering & Chatbots (OpenAI, Hugging Face, Ollama)
- Fine-tuning with LoRA / QLoRA
- RAG (Retrieval-Augmented Generation)
- Embedding Models & Vector Databases (FAISS, Chroma, Weaviate)
- Model Serving (FastAPI, LangChain, Docker, MLflow)
- LLMs used:
LLaMA,Gemma,GPT,Mistral,Phi,DeepSeek,Mixtral,Command-R,BloomZ,Qwen
-
Stable Diffusion (text-to-image, inpainting, img2img, ControlNet, LoRA support)
-
ComfyUI: Custom pipeline design, workflow automation, model chaining
-
Hugging Face Diffusers: Deployment and experimentation with pretrained models
-
DreamBooth & LoRA: Fine-tuning for personalized generation
-
Leonardo.AI & DALL·E 3 API: API-based content generation (posters, ads, etc.)
-
Prompt Engineering for Visual Content: Tailored prompts for marketing, branding and storytelling
-
Post-processing Tools: Seamless pipelines with upscalers, attention masking, aesthetic scoring, and alucination detection
-
Model Management: .safetensors, .ckpt, .onnx model optimization and conversion
- Supervised & Unsupervised Learning
- Regression & Classification
- Time Series Forecasting
- Recommender Systems
- Model Evaluation & Feature Selection
- Artificial Neural Networks (ANN)
- Convolutional Neural Networks (CNN)
- Computer Vision with TensorFlow & PyTorch
- Data Wrangling & Cleaning
- Exploratory Data Analysis (EDA)
- Statistical Analysis & Hypothesis Testing
- Visual Dashboards & Storytelling
- Reporting Automation with Power BI / Streamlit / Tableau
Bringing machine learning models to production with reliability, automation, and scalability.
- Model versioning & registry (MLflow, DVC)
- Experiment tracking and parameter logging
- REST API deployment (FastAPI + Docker)
- Monitoring & Logging (Prometheus, Grafana, Kibana)
- CI/CD pipelines (GitHub Actions, CircleCI)
- Container orchestration (Kubernetes, Docker Compose)
- Workflow automation (Apache Airflow)
- Secure API scaling with NGINX + token-based auth
- Microservices architecture for modular AI services
Building smart edge solutions by combining AI with embedded hardware.
- NVIDIA Jetson Nano – AI edge computing with GPU acceleration
- ESP32 / ESP32-C3 – Low-power microcontrollers with Wi-Fi + BLE
- Arduino – Prototyping with C/C++ and sensor integration
- Raspberry Pi – Full-stack embedded Linux systems for IoT and automation
- Sipeed M1 / M1s Dock – RISC-V microcontroller with AI co-processor (Kendryte K210)
- Real-time sensor data collection
- Edge AI for activity recognition (HAR)
- IoT integration with mobile apps (Flutter + BLE)
- Offline AI inference with optimized models
