Mudassir Ali's Work | ContraWork by Mudassir Ali

Mudassir Ali

Mudassir Ali

AI Chatbot Integration/AI Engineer

New to Contra

Mudassir is ready for their next project!

Rawalpindi, Pakistan

Work Services About

Rawalpindi, Pakistan

Cover image for The Problem: Legal, HR, and

The Problem: Legal, HR, and finance departments are buried under unstructured documents (contracts, resumes, invoices) and struggle to extract key information quickly. The Solution: A specialized AI agent that securely ingests hundreds of pages of unstructured PDFs, extracts specific key-value pairs (like contract expiration dates or invoice totals), and outputs the structured data into a dashboard or spreadsheet. Tech Stack: LlamaIndex, Python, AWS Textract / OpenAI Vision, Streamlit for the dashboard. Key Features Highlight: Optical Character Recognition (OCR) integration. Structured JSON output generation from messy text. Privacy-first architecture (can be showcased using open-source local LLMs like Llama 3).

Cover image for The Problem: Sales teams waste

The Problem: Sales teams waste countless hours researching cold leads, drafting initial outreach emails, and manually updating their CRM. The Solution: An autonomous AI sales agent that scrapes lead information from LinkedIn and company websites, scores the lead based on ideal customer profiles, drafts a highly personalized outreach email, and automatically logs the activity in HubSpot or Salesforce. Tech Stack: CrewAI or AutoGen, Python, Make.com (http://Make.com) / n8n, CRM APIs. Key Features Highlight: Web scraping and data enrichment. Multi-step decision making (qualify -> draft -> update). API integrations with standard CRM platforms

Cover image for The Problem: Traditional chatbots are

The Problem: Traditional chatbots are rigid and fail at answering complex, company-specific questions, frustrating users and escalating too many tickets to human agents. The Solution: An intelligent chatbot integrated with Retrieval-Augmented Generation (RAG). It ingests the company’s FAQs, product manuals, and past support tickets to provide accurate, conversational answers, escalating to a human only when necessary. Tech Stack: Python, LangChain, OpenAI API, Pinecone (Vector Database), React/Next.js for the frontend. Key Features Highlight: Dynamic context retrieval (reads PDFs and website data). Sentiment analysis to detect frustrated customers. Seamless human handoff protocol.