zaryab ahmad

I’m a software engineering professional with a Bachelor of Computer Science from PUCIT Lahore, with coursework focused on software engineering, distributed systems, databases, algorithms, and web technologies. I build full-stack web applications and backend systems through academic and personal projects, with a strong focus on scalable software design and modern development workflows.

At Turing (2025–present), I act as a Software Engineering SME, designing and evaluating AI engineering workflows for backend systems, architecture reviews, and production-readiness scenarios. I evaluated 500+ structured AI prompts to improve engineering accuracy, increasing the consistency of AI-generated implementation outputs by 22% using structured multi-stage evaluation and validation workflows. I also built reusable evaluation rubrics for API security, scalability, and backend architecture reviews, and contributed production-grade coding datasets for supervised fine-tuning pipelines.

Previously, at ARBISOFT in Lahore (2023–2025), I led full-stack development and backend architecture for scalable high-traffic applications. I maintained 99.9% uptime across systems handling 50K+ daily API requests, lowered API response latency by 30% through query optimization and caching strategies, and improved PostgreSQL query performance by 35% through schema redesign and indexing. I strengthened security with RBAC and token-based authentication, reduced unauthorized access incidents, and accelerated delivery by reducing deployment cycles from days to under 1 hour using Docker and CI/CD pipelines.

Earlier at DEVSINC (2021–2023), I developed full-stack web applications and backend services for enterprise platforms supporting internal workflows. I improved internal workflow efficiency by 18%, reduced report generation time from 8s to 5s, and delivered features across 15+ production releases while contributing to system design discussions around APIs, permissions, and reliability. I continue to build ambitious projects—like DEFTGPT and LEARNOS—combining FastAPI/React with RAG pipelines, LLM integrations, and scalable backend APIs, because I enjoy translating complex requirements into production-grade solutions.