I am an AI Engineer specializing in LLMs, RAG systems, and NLP, with hands-on experience designing, building, and deploying end-to-end AI solutions in production. My work spans LLM fine-tuning, hybrid RAG architectures, AI agents, multimodal content generation, and advanced signal processing.
- Design and implement LLM-powered systems (Llama, Gemma, GPT) with hybrid RAG, reranking, and vector search.
- Build AI agents and workflows using LangGraph, LangChain, and agentic architectures.
- Develop multilingual NLP pipelines (classification, NER, sentiment analysis, summarization, translation, embeddings).
- Deploy models to production with FastAPI, Flask, Docker, and CI/CD, focusing on scalability and reliability.
- Combine machine learning, deep learning, and signal processing for real-world, data-driven applications.
-
LLMs & Generative AI
- Fine-tuning and serving Llama, Gemma, GPT-based models
- RAG & Hybrid RAG solutions with reranking and vector databases
- LLM-based classification, routing, and decision systems
- AI agents with LangGraph / LangChain and tool-using workflows
-
NLP & Multilingual Systems
- Semantic search, NER, sentiment analysis, summarization
- Text embeddings (Chroma, Word2Vec, GloVe, BERT, Transformers)
- Multilingual pipelines with adaptive translation, BLEU-based evaluation
- Retrieval and analysis of large document corpora (e.g., academic papers, tourism data)
-
Machine Learning & Signal Processing
- Regression & classification models for forecasting and prediction
- PCA, ICA, Kalman filters, HMMs, Bayesian inference
- Anomaly detection using statistical models and multivariate Gaussian
- Clustering (k-means, EM) and pattern recognition in dynamic environments
-
Languages & Libraries
- Python (PyTorch, NumPy, Pandas, Transformers, Scikit-learn)
- Hugging Face, LangChain, LangGraph
-
LLM & GenAI Ecosystem
- Llama, Gemma, GPT models
- RAG with Chroma, vectorDBs, embeddings
- PEFT, LoRA, diffusion models, Stable Diffusion 3.5
-
Data & Infra
- SQL & NoSQL, PostgreSQL
- Google Cloud Platform, Google Cloud Storage, Google Translate API
- arXiv API, Telegram data pipelines
-
MLOps & Deployment
- FastAPI, Flask, Docker, CI/CD (GitHub/GitLab), REST APIs
- Streamlit for dashboards & web interfaces
-
Tools & Collaboration
- Git, GitHub / GitLab, Jira
- Experiment design, A/B testing, performance monitoring
- Email: sabankara024@gmail.com
- GitHub: @whitemithrandir
- LinkedIn: https://www.linkedin.com/in/sabankara/
- Medium: https://sabankara.medium.com/