Daniyal Asif
Engineering reliable systems for complex workflows.
I architect and deploy focused applications using Python backends, React frontends, and LLM pipelines, optimized for reliability and maintainability.
Current focus
Systems, workflows, delivery
Python, FastAPI, async APIs, deployment
React, Next.js, Streamlit, practical UX
Tool-calling, prompt workflows, inference boundaries
Transparent outputs, failure handling, provenance
Selected projects that show current direction and strong delivery.
Profile
A builder with a systems-first mindset.
I work across the Full-stack: backend services, responsive interfaces, LLM-integrated workflows, and deployment. The common thread is building useful systems with clear behavior, visible limits, and maintainable architecture.
Recent work includes a genomic variant review copilot with explicit safety boundaries, a real-time fleet coordination simulation shipped under hackathon constraints, and a deployed prompt enhancement tool.
- Solo delivery of a 2,100+ LOC real-time simulation engine in 48 hours.
- Experience with model-backed workflows where errors and uncertainty must be visible.
- Hands-on ownership from architecture and UI through cloud deployment.
Connect
Credentials
Location
Lahore, Pakistan
Education
Coursera (Google Professional Certificate)
Information Technology (2025)
freeCodeCamp
Information Technology (2022-2024)
Certifications
- Google IT Automation with Python
- Google Advanced Data Analytics
- Google AI Essentials
- Google Prompting Essentials
- AWS AI Practitioner Challenge
- LabLab.ai Launch Fund Hackathon
Languages
English (Full Professional), Urdu (Native), Hindi (Native)
Capabilities
Stack for building, shipping, and explaining systems.
Work
Only the projects that best represent the current state of my work and thinking.
A Streamlit and Python system for evidence-led variant review, designed around transparent outputs, provenance, and explicit product boundaries in a sensitive domain.
- Decoupled evidence lookup from model scoring to keep review outputs auditable.
- Implemented visible LIVE / DEMO modes with clear provenance and safe fallback behavior.
- Handled inference failures explicitly instead of hiding uncertainty behind generic responses.
A disaster-response fleet simulation built within a 48-hour hackathon sprint to simulate autonomous routing logic, combining real-time coordination, pathfinding, and a clear operator dashboard.
- Built a 3-tier FastAPI and Next.js system with more than 2,100 lines delivered under hackathon constraints.
- Implemented 1-second SSE refresh cycles for live fleet synchronization.
- Engineered A* pathfinding over a 2,500-cell grid with obstacle-aware routing.
A deployed developer tool that turns rough prompts into clearer, structured instructions for LLM workflows, with a simple interface and modular production deployment.
- Designed a practical refinement workflow for clearer LLM instructions and evaluation.
- Kept the architecture modular so prompt logic, UI, and deployment concerns remain separate.
- Maintained the live Vercel deployment end to end.