Muhammad Imran
Software Engineer · AI · Web · Open Source
Welcome to my universe.Where ideas become systems that run in the real world.
Hi, I’m Muhammad — a software engineer through and through. I translate vague human desires into strictly logical machine tantrums.
Proof before promises
Selected Work
Open source, AI, and full-stack products with clear ownership, measurable outcomes, and working evidence.
Zowe Client Java SDK
Contributing production Java methods that help enterprise developers automate IBM z/OS mainframe workflows.
- Role
- Linux Foundation LFX Mentee
- Built
- 11 z/OSMF workflow API operations
- Verified
- Open-source review and Maven delivery
FitGPT
Built the backend for an AI wardrobe assistant that turns real wardrobe, weather, and trip data into useful recommendations.
- Role
- Backend Developer across 3 Agile sprints
- Built
- 12+ product features and integrations
- Verified
- 185+ pytest tests and GitHub Actions CI
SolarShare
Co-created a solar-energy marketplace with auditable savings, payment rails, and role-based workflows.
- Role
- Co-Founder and Full-Stack Developer
- Built
- 3-role RBAC and 4-stage billing lifecycle
- Verified
- 1st place, 2025 PSEGLI Challenge
Experience
Where I’ve worked
Enterprise support, entrepreneurship, open source, backend, mobile, mentoring, and machine learning.
Software Support Technician
- Supporting enterprise grocery retail and wholesale software across 14+ modules and client-facing workflows.
- Resolving issues through tickets, remote troubleshooting, release QA, and reusable documentation.
LFX Mentee — Zowe Client Java SDK
- Implementing 11 Java SDK methods for the complete z/OSMF Workflow REST API lifecycle.
- Contributing reviewed open-source code used by enterprise developers automating IBM z/OS mainframes.
Java and Flutter Developer Intern
- Shipped 10+ Java REST APIs and a Flutter application released for iOS and Android.
- Worked with technical leadership through scoped sprints, code review, debugging, and release validation.
AI4ALL Ignite Fellow
- Built an end-to-end machine-learning pipeline from raw data through feature engineering and evaluation.
- Presented metrics, trade-offs, and reproducible results to a technical panel.
How I work
Clear engineering, end to end
Trace the actual failure path before touching code. Fix the cause, not the symptom.
Design clear boundaries so features remain understandable, testable, and safe to extend.
Ship across UI, API, and data with verification that holds up after the sprint ends.
Applied AI
AI, used like an engineer.
I completed Anthropic Academy’s AI Fluency and MCP Advanced Topics, finished LaunchCode AI 101 by building a measured email-triage automation, and was accepted into the Agentic Engineer track. I also shipped Groq and Llama 3.1 recommendations in FitGPT behind 185+ tests and built an end-to-end machine-learning pipeline through AI4ALL. AI accelerates the work; review, testing, and ownership still decide what ships.
Next step