MSc Artificial Intelligence graduate & Agentic AI Engineer building production backend microservices (Python, FastAPI, Ollama, ChromaDB, PyTorch) deployed on cloud infrastructure (AWS, Azure, Docker). Based in Glasgow, UK.
kalki@agentic-core:~$ systemctl status agentic-pipeline
● agentic-pipeline.service - Active (running) since May 2026
kalki@agentic-core:~$ cat core_capabilities.json
kalki@agentic-core:~$ Type 'help', 'skills', 'projects', or 'clear'..._
I am an Agentic AI Software Engineer and MSc Artificial Intelligence graduate with hands-on experience designing, building, and deploying Generative AI and RAG-based applications as production backend services using Python, FastAPI, and cloud-native infrastructure (Azure, AWS, Docker).
My expertise spans LLM API integration (OpenAI, Anthropic, Ollama), vector search & embeddings (ChromaDB, sentence-transformers), prompt engineering, and CI/CD delivery practices. I hold a solid foundation in PyTorch, TensorFlow, and structured model evaluation (precision, recall, F1, ROC AUC, GridSearchCV). Comfortable operating end-to-end—from data pipeline design through to production integration, testing, documentation, and stakeholder delivery in regulated commercial environments.
MSc in Artificial Intelligence
Glasgow, United Kingdom
Agentic LLMs & Production RAG Systems
Agentic LLM workflow design, RAG pipeline design (ChromaDB, sentence-transformer embeddings, vector search), LLM/SLM integration (OpenAI, Anthropic, Ollama), model serving, prompt engineering, validation.
Python, FastAPI, RESTful API design, SQL, Git, AI-assisted development practices with code review and automated testing.
Azure, AWS (model training & deployment), Docker, CI/CD pipelines, MLOps tooling, end-to-end data pipeline engineering.
PyTorch, TensorFlow, Scikit-learn, model evaluation frameworks (precision, recall, F1, ROC AUC), hyperparameter tuning (GridSearchCV).
Stakeholder engagement & technical communication, methodology documentation, moving solutions from prototype to stable production readiness.
Engineered applications spanning Generative AI, RAG, NLP, IoT networks, and Computer Vision.
Built a retrieval-augmented document question-answering application combining a FastAPI backend microservice, TF-IDF based retrieval, and LLM APIs to return grounded answers directly from source documents.
Designed and fine-tuned transformer models (BERT embeddings) combining classical and neural information retrieval techniques into an agentic search system, improving semantic query relevance by 28% (MAP score).
Built and evaluated time-series predictive ML models end-to-end on AWS, covering feature engineering, model selection, training, and validation on sequential traffic data—achieving 35% higher throughput and 20% lower latency.
Designed and trained a hybrid CNN and SVM pipeline for medical image classification in MATLAB, covering image preprocessing, feature extraction, model training, and evaluation against ground truth labels for medical cyst classification.
Designed an end-to-end Python data pipeline (Scikit-learn) with feature scaling and stratified train/test splitting, training Logistic Regression and Random Forest classifiers using GridSearchCV, ROC AUC, precision, recall, and F1 score.
MSc in Artificial Intelligence
Relevant Modules: Ethical AI, Machine Learning, Computer Vision
BEng in Computer Science
Available for AI Engineering, Agentic Backend, and MLOps roles.