Problem: AI advisory systems need controlled personas, structured knowledge, configurable workflows, and safer AI responses instead of generic chatbot behavior.
Solution: Built a configurable AI Advisor System with admin-managed personas, RAG-based knowledge flow, multi-agent workflows, LiteLLM integration, Vertex AI services, embeddings, and cloud-backed AI orchestration.
Impact: Created a scalable AI advisory foundation capable of supporting controlled multi-persona assistant workflows with structured knowledge retrieval and cloud AI integrations.
Problem: Recruiters and clients need a faster way to understand a candidate’s skills, experience, projects, and role fit without reading the full portfolio manually.
Solution: Built an AI-powered portfolio assistant with public chat, admin-managed profile data, AI instructions, knowledge base, API keys, projects, skills, experience, certifications, and social links.
Impact: Created an interactive portfolio experience that helps HRs, clients, and technical reviewers understand the candidate profile through natural questions.
Problem: Recruitment teams need faster resume screening, JD matching, candidate ranking, and recruiter-ready summaries without losing structured evaluation context.
Solution: Built an API-first AI recruitment assistant for resume analysis, job-description matching, profile ranking, chat-based evaluation, and analytics-supported shortlisting.
Impact: Created a repeatable AI-assisted screening workflow that can help recruiters compare candidates faster and make more structured shortlisting decisions.
Problem: Electronics inventory operations required better product tracking, stock visibility, order handling, and inventory management through a centralized system.
Solution: Built a MERN stack-based Inventory Management System for managing products, stock updates, inventory workflows, and admin operations through a web-based dashboard.
Problem: Climate data needed to be organized, visualized, and modeled to make warming trends easier to understand and compare.
Solution: Developed a JavaFX desktop application backed by MySQL and JDBC with multivariate linear regression for climate trend analysis and prediction.
JavaJavaFXMySQLJDBCLinear Regression
Impact: Turned historical climate records into a more approachable analysis, visualization, and prediction workflow.
Featured