Ashhad Ahmed Kamran
Building Adaptive Agentic Systems
AI Engineer · GPA 3.93/4.0 · Ranked 1st · Dawood University
Professional Summary
Results-driven AI Engineer with couple years of experience in software development using Python. Specialized in building adaptive agentic AI systems, prompt chaining, and programmatic feedback loops with human-in-the-loop integration. Expert in RAG architectures, data manipulation, and deploying evaluation agents to ensure LLM reliability. Proven ability to translate complex machine learning concepts into scalable production solutions.
Education
BS in Artificial Intelligence
Dawood University of Engineering and Technology | Karachi, Pakistan
Core Focus: Python, DSA, AI/ML, Neural Networks, Predictive Analytics, Data Science.
GPA: 3.93/4.0 · Ranked 1st in Batch
Exchange Semester: Applied CS and AI
Sapienza Università di Roma | Rome, Italy
Relevant Coursework: Advanced Machine Learning, Agentic AI, Computer Vision Lab.
Erasmus+ Exchange Program Scholar
Work Experience
AI Engineer (Python Developer)
Onestop Vendors | Karachi, Pakistan
- Architected adaptive agentic AI systems using LangGraph, utilizing prompt chaining to automate complex customer workflows, improving response relevance by 40%.
- Developed programmatic feedback loops with human-in-the-loop checkpoints for high-stakes voice/chat interactions, ensuring 100% accuracy in data capture.
- Integrated Qdrant/FAISS vector databases and optimized data manipulation scripts, reducing system latency by 30%.
- Engineered real-time STT/TTS pipelines, reducing interaction delays by 500ms for enterprise-grade AI agents.
AI Intern
Arcana Info (Pvt.) Ltd | Karachi, Pakistan
- Deployed ML models on IBM Watson using Scikit-learn for predictive intent analysis, achieving 92% accuracy.
- Automated complex data reporting using Python and IBM Analytics, saving 10+ hours of manual work weekly.
Technical Skills
Technologies and tools I work with to build intelligent systems.
Machine Learning
AI Systems
Programming Languages
Tools & Cloud
Projects
A showcase of AI & ML engineering work, from research to production systems.
EmCon AgenticAI Based Communication Tool
Autonomous AI agent using prompt chaining to manage scheduling via email, with evaluation agents for calendar verification and human-in-the-loop confirmations.
Key Achievements
- Reduced manual scheduling time by 95%
- Human-in-the-loop checkpoint integration
- Programmatic feedback loop architecture
Multimodal Visual Search Engine
End-to-end image retrieval system using ResNet18 and GRU with PCA dimensionality reduction on a 10k+ image dataset. Features semantic search and top-K retrieval.
Key Achievements
- 94% top-5 accuracy on image retrieval
- Processed 10,000+ image dataset
- Streamlit UI with real-time search
AI-Powered Voice & Sales Automation
Enterprise-grade voice agent with IVR system leveraging LLM reasoning to intelligently route customer intents. Features real-time STT/TTS with sub-500ms latency.
Key Achievements
- Routes 98% of customer intents correctly
- Reduced interaction delays by 500ms
- Edge case handling via prompt engineering