Computer Science Graduate · ML & AI — Lahore, PK
Computer Science graduate and AI Automation Engineer building machine learning, NLP, and LLM applications — RAG pipelines, fine-tuned models, and automation workflows — currently a Business Development Associate at Aslase after an AI automation internship there, open to roles in Germany, and co-author of two published AI research papers.
About
I'm a Computer Science graduate from the University of Management & Technology, Lahore (BSCS, 2026), working across AI and software engineering. My focus is building practical systems in AI automation, machine learning, NLP, LLM applications, and RAG — not just training models, but connecting them into workflows that run end-to-end. I'm currently a Business Development Associate at Aslase, after an AI automation internship there building n8n and LLM workflows, so I work on both the technical and the client side of AI projects. Alongside AI work I have hands-on software, web, and mobile development experience, and I'm the co-author of two published research papers on multi-agent LLM systems and data-centric AI.
Experience
Qualify inbound and outbound leads and run discovery calls for AI, automation and software projects. Coordinate project delivery between clients and the engineering team, and turn client problems into AI solution proposals, drawing on hands-on automation experience.
Built and maintained automation workflows in n8n connecting LLMs, APIs and business tools, wrote Python scripts and API integrations that removed repetitive manual steps, and contributed to an AI marketing intelligence product.
Contributed to mobile application development across the software development lifecycle — building and testing application features, debugging, and collaborating with the team to improve application performance and user experience.
Developed and maintained responsive web pages and UI components using HTML, CSS, and JavaScript. Collaborated with design and backend teams to implement UI components and fix bugs, and tested across browsers and devices to ensure consistent performance and a strong user experience.
Research & Publications
Co-authored research bridging applied AI engineering and academic work.
Co-authored research on an LLM-based multi-agent system that uses the Model Context Protocol (MCP) and Agent-to-Agent (A2A) protocol to coordinate specialized agents for emergency-response workflows — covering agent discovery via agent cards, tool and resource interaction, and agent evaluation using LLM-as-a-Judge and BERTScore.
Co-authored a study on data-centric AI showing how quality-driven data engineering — systematic cleaning, annotation, augmentation, and validation — improves model accuracy, generalization, and fairness across machine learning models, and examining data governance within AI pipelines.
Projects
Ordered to lead with flagship AI/ML builds, then model depth, then applied pipelines.
A production-style AI mock interview platform that parses a candidate's CV, generates personalized questions with a Qwen2.5 LoRA model, delivers them via Piper TTS through an avatar, transcribes spoken answers with local Whisper, and scores responses into a structured feedback report.
An end-to-end kidney CT analysis system: a deep-learning classifier labels scans as Cyst, Normal, Stone, or Tumor, a 2D U-Net with 2.5D neighboring-slice input segments tumor regions, and the outputs drive explainability, report generation, and 3D/VR visualization.
An AI marketing and sales intelligence product at Aslase that brings marketing and analytics data together in dashboards and adds AI-generated insights on top. I contributed during my internship.
A document-based question-answering system using LangChain and ChromaDB — document loading, chunking, sentence-transformer embeddings, and vector retrieval feeding LLM APIs for answers grounded in retrieved context.
Fine-tuned a pretrained language model on a medical Q&A dataset using QLoRA — a memory-efficient approach that made training feasible on limited GPU resources, with training and evaluation via Hugging Face.
A pipeline combining Whisper speech-to-text with a language model to generate automated responses, integrating multiple AI components into a single end-to-end workflow.
Stack
Certifications