Lakshitha Vimuth

AI/ML Integration in IoT:

✔ Built real-time AI pipelines for IoT sensor data analytics, reducing equipment failures by 40%.

✔ Integrated Hugging Face models for anomaly detection in industrial networks.

Data-Driven IoT Analytics:

✔ Applied ML models for real-time IoT sensor data analysis, enhancing decision-making and predictive maintenance.

Generative AI for Business Automation:

✔ Developed AI-powered chatbots using LangChain + OpenAI to automate customer support.

✔ Designed AI-driven workflow automation, reducing manual workload by 60%.

Custom IoT Solutions:

✔ Directed on-site evaluations and crafted AI-powered IoT solutions tailored to client specifications, boosting operational efficiency and achieving a 120% surge in market demand.

Innovation and Scalability:

✔ Developed scalable AI-integrated IoT frameworks, accommodating increased data and complexity, and spearheaded continuous tech advancements.

MLOps & AI Model Deployment:

✔ Deployed AI models on AWS, Azure, using FastAPI & TensorFlow Serving.

✔ Built CI/CD pipelines for continuous AI model updates and monitoring.

Research and Development:

✔ Fine-tuned BERT, BioBERT, and GPT-based models for medical document summarization.

✔ Developed NLP-based clinical decision-support systems for automated diagnosis.

Experimental Design and Analysis:

✔ Designed and executed complex experiments, employing statistical analysis and ML algorithms to validate hypotheses and interpret data, driving forward scientific understanding.

Technical Demonstrations:

✔ Led practical demonstrations and workshops for AI and ML, simplifying complex concepts for students and enhancing their hands-on skills in Python, Java, R, and C++.

AI for Medical Imaging & Diagnostics:

✔ Created AI-powered lung cancer detection models achieving 98% accuracy.

✔ Applied GANs & Transformers for CT scan image enhancement & segmentation.

Educational Support:

✔ Provided comprehensive academic support to students, including tutoring, feedback on assignments, and guidance on project work, ensuring a deep understanding of subject matter and application.

Innovative Teaching Methods:

✔ Developed and implemented innovative teaching methods and materials, including interactive digital content and simulation exercises, to engage students and improve learning outcomes.

Laboratory Management:

✔ Oversaw the maintenance and operation of research and teaching laboratories, ensuring the availability of up-to-date equipment and compliance with safety standards.

Conference Participation and Networking:

✔ Actively participated in academic conferences and seminars, presenting findings and networking with peers, contributing to a vibrant academic community, and staying abreast of industry trends.

AI Optimization for Edge Devices:

✔ Built lightweight ML models for low-power edge computing, increasing efficiency by 500%. ✔ Used TensorRT & ONNX to optimize AI inference on IoT & embedded devices.

LLMs & Conversational AI:

✔ Experimented with GPT-based models for voice assistants and customer support chatbots.

AI Model Deployment & Scaling:

✔ Deployed ML models using Docker, Kubernetes, and AWS Lambda, ensuring scalability and cost efficiency.

The project can involve the development of a computer vision system that uses machine learning algorithms and image processing techniques to identify medicinal plants and detect fungal diseases on their leaves. The system can be trained on a dataset of images of different medicinal plants and their corresponding fungal diseases to learn the visual patterns and characteristics of each.

The project can involve the development of a machine learning model that uses temporal and meteorological data to predict the Air Quality Index (AQI) for a particular location. The project can also involve the use of advanced machine learning algorithms such as Long Short-Term Memory (LSTM) networks and Gradient Boosting Machines (GBMs) to improve the accuracy of the prediction and handle the complexity of temporal data.

The project was done as a mini-project at the University. The project aimed to replace the manual gate system with an automated one. A license plate recognition system was developed using YOLO and CNN. The system was tested with a small-scale prototype. Used YOLO and CNN to detect and recognize license plates.Developed a small-scale prototype to simulate the process. Developed a Flask web app on Heroku to interact with the system.

The project can involve the creation of a web or mobile application that enables organizations to track the attendance of their employees or students in real-time. The application can include features such as automatic marking of attendance based on location or time, real-time monitoring of attendance records, and integration with payroll or academic systems.