Specializing in cutting-edge artificial intelligence at Carnegie Mellon University's Information Networking Institute. With expertise in building intelligent systems, along with knowledge of Deep Neural Networks, Generative AI, and Large Language Models.
I'm Prajwal Kumar, AI Engineering Leader currently pursuing my MS in AI Engineering at Carnegie Mellon University. Being part of CMU's AI community has allowed me to collaborate with leading researchers while developing ethical AI systems across a diverse range of projects from medical image classification to language identification models and multilingual RAG pipelines.
My research has been published in IEEE, focusing on integrating deep learning techniques for healthcare. As Internal VP of CMU's AI Club and Vice President of GOINI student group, I've collaborated with faculty researchers on adversarial machine learning projects.
I have experience in developing cutting-edge AI solutions with 95%+ accuracy rates, utilizing advanced technologies like TensorFlow, PyTorch, LangChain, and various cloud platforms. I'm passionate about solving real-world problems and ultimately rendering better services to customers.
Specializing in Deep learning, generative AI, and LLMs
Building scalable data pipelines and systems
IEEE published researcher in AI and healthcare
Leading cross-functional teams on AI initiatives
My professional journey in AI/ML and Data Science
Engineered an agentic AI system using LangChain and Playwright to fully automate the customer phone purchase workflow on telecom websites, enabling robust, end-to-end anomaly detection.
Engineered medicine prescription ontology using TensorFlow, email phishing detection with 96% accuracy, AI chatbot for mental health, and more.
Organized hackathons and workshops on GenAI, developed projects on Intel® DevMesh, participated in Intel® AI Global Impact Festival.
Built driver drowsiness detection with 94% accuracy, NLP-based chatbot, and image classification using CNN and SE Networks.
Developed RASA chatbot, generated synthetic fingerprint data using GAN, built language identification using Bidirectional LSTM with 98% accuracy.
Pursuing MS in AI Engineering - Information Security at CMU's Information Networking Institute. Focusing on Deep Learning and AI/ML production with Information Security.
Completed undergraduate studies in computer science with focus on machine learning and data science.
A showcase of my technical work and research
Built an end-to-end data pipeline using SparkML to preprocess, model, and predict player ratings from the FIFA 22 dataset, leveraging distributed computing for large-scale sports analytics.
Designed a multi-modal malware detection pipeline using CNN (ResNeXt), XGBoost (n-grams), and LLaMA-based LLMs, achieving 99.68\% image classification accuracy and building a real-time GUI with JSON-based LLM explainability.
Built a scalable KNN-based movie recommender (NDCG@10 = 0.9983, 65K+ QPS) with Dockerized microservices, Kubernetes, CI/CD (Jenkins), and CRON-based retraining; integrated A/B testing and real-time monitoring with Grafana.
Built a multimodal AI companion with speech-to-text (Google API), CNN-based SER (71.2\%), and Transformer models; improved LLM empathy by 30\% via psychoanalysis modules and deployed a Streamlit app for real-time emotion feedback.
Research on advancing image security through deep learning and cryptography in healthcare, published in IEEE, integrating advanced security techniques with machine learning.
Technical proficiencies and areas of expertise that I've developed
Let's connect and discuss potential collaborations
prajwalk@andrew.cmu.edu
+1 (412) 728-2613
Pittsburgh, Pennsylvania, United States