Projects using OpenAI in LahoreProjects using OpenAI in Lahore
Projects
People
Cover image for Echo is a multi-tenant SaaS
Echo is a multi-tenant SaaS for AI customer support with realtime chat, an AI voice agent, and AI automation. I was the primary developer on the full build, frontend, backend, and the AI layer. The AI is the core. It runs OpenAI, Claude, Gemini, and Grok through one multi-model setup, so a client can switch providers without a rewrite. A RAG pipeline connected to a vector database grounds every answer in the client's own content, so the chatbot never returns generic output. VAPI powers the voice agent, so customers can speak to support on a live call. Each tenant gets its own AI agent built on its own documents. The stack is Next.js 15 and React 19 inside a Turborepo monorepo, with separate apps for the dashboard, the embeddable chat widget, and backend services. Realtime chat runs on Convex. Clerk handles auth. API keys are encrypted per tenant through AWS Secrets Manager, so no two clients share credentials or data. Launch day held 60 live conversations at once with zero dropped sessions. What the client gets: an AI chatbot and voice agent that answer from their own content, work across multiple LLM providers, and stay isolated and secure per tenant.:
2
2
125
Cover image for Project Title: Engineering an Automated
Project Title: Engineering an Automated Intelligence Pipeline (RAG Workflow) The Challenge: In a world drowning in data, the problem isn't having information; it's accessing it instantly. Most businesses leave their most valuable assets buried in Google Drive folders where they remain static and underutilized. As a CS student and designer, I saw an opportunity to bridge the gap between "stored data" and "active intelligence." I wanted to build a system that doesn't just store files but understands them. The Solution: I engineered a sophisticated Retrieval-Augmented Generation (RAG) pipeline using n8n for orchestration. This isn't just a simple link-up; it’s a multi-layered architectural approach to data automation: Data Ingestion: Automatically triggering on Google Drive updates to ensure the AI's "brain" is always current. Vector Processing: Implementing a Recursive Character Text Splitter to ensure data chunks maintain semantic meaning. Memory & Storage: Utilizing Pinecone Vector Store for high-speed similarity searches, allowing for near-instant retrieval of relevant context. Model Integration: Leveraging the power of Google Gemini as the reasoning engine, supported by OpenAI Embeddings for high-precision vectorization. The Intersection of Design & Code: Being a 4th-semester BSCS student, I focused heavily on the logic of the workflow—ensuring the recursive splitting didn't lose context and that the vector database was optimized for performance. However, my background as a Graphic Designer allows me to visualize these complex backend processes into a user experience that feels intuitive. I believe that an automation is only as good as its usability; if the interface is clunky, the power of the AI is lost. Why This Matters for Your Business: This workflow effectively creates a "Custom Brain" for your organization. Imagine chatting with your company’s entire history, SOPs, and technical documents as if you were talking to an expert teammate. By automating this pipeline, I eliminate the need for manual data entry or tedious searching, allowing teams to focus on high-level creative work while the AI handles the information retrieval. Technical Stack Used: Logic: n8n Workflow Automation LLMs: Google Gemini & OpenAI Database: Pinecone (Vector Search) Storage: Google Drive API Theory: RAG (Retrieval-Augmented Generation) & Semantic Search
1
25