Pranesh Vittal Kalikar

Results-oriented AI-Driven Software Engineer with expertise in automation, API testing, and cloud-native systems. Skilled in Python, CI/CD, and validation frameworks for data-driven applications. Experienced in LLMs, prompt engineering, Kafka, Kubernetes, and AWS/Azure. Data Science Intern with a background in data analysis, machine learning, and predictive modeling, driving efficiency improvements through automation.

• ITIL 4 Foundation
• Cisco CCNA
• Certified Python Developer

• Distributed Image Classification System, PyTorch, PySpark, Hadoop, Google Colab, 05/24, 07/24, AI Developer, Designed a distributed CNN-based image classification system using PyTorch with GPU-accelerated training on Google Colab, achieving 92% accuracy across a multi-stage gender and age-group prediction pipeline., Processed 500K+ images using PySpark on a Hadoop cluster, reducing model training time by 55% through distributed data parallelism — enabling faster iteration cycles and improved scalability for large-scale datasets., Architected a multi-stage classification pipeline (gender → age group) that improved model modularity, making it easier to retrain individual stages independently and extend the system to new classification tasks.
• Phishing URL Detection using Transformers (URLTran Model), PyTorch, Transformers, NLP, Scikit-learn, Academic Project, Built an end-to-end NLP pipeline using PyTorch Transformers with tokenization, MLM pre-training, and task-specific fine-tuning, achieving 97% detection accuracy with high ROC-AUC scores against TF-IDF and CNN baselines., Benchmarked Transformer, CNN, and TF-IDF architectures to quantify accuracy trade-offs, producing comprehensive precision, recall, and F1-score reports that guided model selection and justified the Transformer approach.
• Green Grocer System – Full Stack Application, Full Stack Developer, 5, Developed a full-stack grocery inventory and order management application using Agile methodologies, delivering product browsing, ordering, and automated out-of-stock alert features that improved inventory visibility., Contributed to system design and requirements gathering in a 5-person Agile team, enabling cleaner feature scoping and reducing rework by aligning on architecture decisions early in the sprint cycle.
• School Management System – Attendance Automation, Automation Test Engineer, Developed Python-based automation scripts for attendance capture workflows and business rule validation (e.g.,