← Back to Portfolio
AI Models

Agents, LLMs & ML Systems

Agentic AI systems, LLM-powered pipelines, NLP automation, and machine learning models built for real-world business automation.

Work in Progress: My AI & Machine Learning Final Year Project (FYP) is currently under active development. This section will soon be updated with neural network architectures, model evaluation metrics, and live predictive demos.
AGENT ORCHESTRATOR LLM Core INPUT AGENT SQL AGENT POWER BI AGENT OUTPUT AGENT ML PIPELINE scikit-learn · Python parse input query db render viz send report LLM BACKBONE GPT-4o · LangChain
Python LangChain GPT-4o SQL Server Power BI ML
Retail SME Automation — Agentic AI System

Final Year Project — an autonomous multi-agent AI pipeline for retail SME operations. An LLM orchestrator delegates tasks to specialized sub-agents: input parsing, SQL querying, ML inference, Power BI report generation, and output delivery. Reduces manual data work by ~80%.

● In Progress — FYP View GitHub →
INPUT DOC NLP ENGINE spaCy · BERT LLM Extraction STRUCTURED Entity: "Invoice #4821" Amount: "Rs 48,200" Date: "2025-11-14" Vendor: "Tech Corp" Category: "Software" → SQL INSERT ready Unstructured PDF/Text → Structured DB Records
Python NLP LLMs spaCy SQL Server
NLP Document Processing Pipeline

An NLP-driven pipeline that ingests unstructured documents (invoices, reports, emails) and uses LLM extraction to parse, categorize, and insert structured records into SQL Server — eliminating manual data entry workflows.

● Concept / In Dev View GitHub →