class Kumaran:
def __init__(self):
self.role = "AI/ML Engineer"
self.identity = "Tech enthusiast who builds systems, not just models"
self.domains = [
"Artificial Intelligence",
"Machine Learning",
"Data Science",
"Generative AI",
"Agentic AI Systems"
]
self.current_focus = [
"Designing end-to-end AI and ML pipelines",
"Building production-ready GenAI and agentic workflows",
"Applying retrieval, reasoning, and evaluation in real systems"
]
self.tools_and_systems = [
"LLMs, RAG pipelines, and vector search",
"Conversational and speech-based AI systems",
"Scalable APIs and deployment-ready architectures"
]
self.principles = [
"Production over prototypes",
"Systems thinking over isolated solutions",
"Learning through building and iteration"
]
def fun_fact(self):
return "I enjoy working on AI systems where reliability and real-world performance actually matter."
me = Kumaran()
- Bengaluru, India
-
21:48
(UTC -12:00) - in/kumaran-elumalai
Pinned Loading
-
ai-powered-financial-news-intelligence-system-using-langgraph
ai-powered-financial-news-intelligence-system-using-langgraph PublicAI-powered financial news intelligence system built with LangGraph multi-agent pipelines, vector embeddings, strict impact ranking, and local LLMs for extractive reasoning, enabling high-precision …
Jupyter Notebook 1
-
nextgen-multimodal-generative-vlm-evaluation-suite
nextgen-multimodal-generative-vlm-evaluation-suite PublicA benchmark suite for lightweight generative multimodal Vision-Language Models, comparing ViLT and SmolVLM under resource-constrained inference environments. Demonstrates CPU-only deployment, model…
Python 1
-
building-RAG-agents-with-LLMs-NVIDIA
building-RAG-agents-with-LLMs-NVIDIA PublicEnd-to-end implementation of Retrieval-Augmented Generation (RAG) agents with LLMs using NVIDIA AI Endpoints, a hands-on course project demonstrating vector search, prompt engineering, and evaluati…
Jupyter Notebook
-
hotel-sentinel
hotel-sentinel PublicHotelSentinel is an NLP-driven sentiment intelligence system that analyzes hotel reviews to classify customer sentiment and uncover key experience drivers. It combines text preprocessing, TF-IDF fe…
Jupyter Notebook
-
persona-sense
persona-sense PublicAn end-to-end machine learning system designed for customer intelligence and persona discovery. It leverages unsupervised learning to segment customers based on behavioral and demographic patterns,…
Jupyter Notebook
-
aurum-forecast
aurum-forecast PublicAurumForecast is an end-to-end time-series forecasting system designed to predict gold prices over a 30-day horizon. It evaluates multiple statistical and econometric models to capture trend and se…
Jupyter Notebook
If the problem persists, check the GitHub status page or contact support.
