Developed an AI-powered FAQ chatbot automation system using n8n, Google Sheets, and OpenAI to provide automated, contextual, and real-time responses to user queries through a website chatbot interface. The project uses a RAG (Retrieval-Augmented Generation) architecture where chatbot responses are dynamically generated based on FAQ data stored in a single Google Sheet. The system retrieves the most relevant FAQ information and augments AI-generated responses for accurate and human-like customer support interactions. Key Objectives Automate repetitive customer support queries Reduce manual response handling Create a scalable AI FAQ assistant Enable non-technical teams to manage FAQs easily Integrate chatbot with website frontend

**Linux-Based AI Agent for Intelligent System Administration** is an AI-powered web application developed using Python, Flask, Linux system utilities, and the OpenAI API to simplify Linux system administration through natural language interaction. The project provides an intelligent AI terminal assistant where users can execute Linux commands, monitor system processes, retrieve hardware information, and receive explanations for complex terminal commands using simple conversational queries instead of memorizing difficult command-line syntax. Traditional Linux administration can be challenging for beginners and time-consuming for administrators due to the need for extensive command knowledge, troubleshooting skills, and manual operations. This project addresses these issues by integrating artificial intelligence with Linux automation, enabling users to interact with the system more efficiently and intuitively. The AI agent improves productivity by reducing manual effort, assisting in command understanding, and automating administrative tasks through a user-friendly web interface. The system acts as a bridge between human language and Linux system operations, making Linux management more accessible, interactive, and efficient for both learning and real-world administration tasks.

1) RAG Pipeline & Chatbot ## First I choose a trigger name is Google Drive Trigger and connected to the https://console.cloud.google.com/ and enable google API Key. ## Second I attached a seond trigger which is Download File, and attach json ID ## Third Create an Pinecone Database connect with the nodes and generate API key and create index that connected with the Embedding Open AI Model that helps me to chunk the data.

1) RAG Pipeline & Chatbot ## First I choose a trigger name is Google Drive Trigger and connected to the https://console.cloud.google.com/ and enable google API Key. ## Second I attached a seond trigger which is Download File, and attach json ID ## Third Create an Pinecone Database connect with the nodes and generate API key and create index that connected with the Embedding Open AI Model that helps me to chunk the data.

1) RAG Pipeline & Chatbot ## First I choose a trigger name is Google Drive Trigger and connected to the https://console.cloud.google.com/ and enable google API Key. ## Second I attached a seond trigger which is Download File, and attach json ID ## Third Create an Pinecone Database connect with the nodes and generate API key and create index that connected with the Embedding Open AI Model that helps me to chunk the data.

The application compares user's uploaded resume with the given job description and gives the ATS matching score. It displays a score, the candidate's strengths and missing skills and tools.

which team can make goal fast

which team can make goal fast

dvwc

An application that helps people to have basic knowledge of investment. As peopl are starting to invest more but still making loss. So used ai to make working ai tool for this and prototype.

sdvsdc

I built an AI agent that decides loan eligibility in under 2 seconds — and explains every decision in plain English. Here's the problem it solves (and why it matters): ───────────────────────── 𝗧𝗵𝗲 𝗣𝗿𝗼𝗯𝗹𝗲𝗺 Loan approvals in traditional banking are: ❌ Slow — taking days or even weeks for a basic eligibility check ❌ Inconsistent — different officers, different outcomes for the same applicant ❌ Opaque — applicants get rejected with zero explanation ❌ Disconnected from policy — credit scores rarely reflect the latest RBI guidelines In a world where speed and fairness define customer trust, this is a broken experience. ───────────────────────── 𝗧𝗵𝗲 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻: 𝗟𝗼𝗮𝗻 𝗘𝗹𝗶𝗴𝗶𝗯𝗶𝗹𝗶𝘁𝘆 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 I built an end-to-end AI-powered agent that evaluates loan applications in a fraction of the time — with full transparency. Here's how it works: ✅ Applicant submits details via a clean Streamlit web interface ✅ A scoring engine instantly calculates EMI, debt ratios, and hard eligibility rules ✅ An RAG module retrieves the latest RBI guidelines relevant to the application ✅ A Llama 4 Scout LLM (via Groq) generates a human-readable explanation ✅ The agent outputs a clear Approve or Reject decision — with reasoning Every step is orchestrated through a LangGraph state-graph pipeline, making the workflow fully auditable and deterministic. 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗮𝘀 𝗮 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 As a PM, what excited me most wasn't the tech — it was the outcome: → Decision time drops from days to under 2 seconds → Every rejection comes with a reason the applicant can understand → Compliance is baked in, not bolted on — RBI guidelines retrieved per query → If the LLM ever fails, a rule-based fallback kicks in — zero downtime This is what AI-first product design looks like in regulated industries.