Practical lessons from building native AI products, leading delivery, and keeping systems secure.
I didn’t “switch careers” into prompt engineering. I layered it on top of years of engineering and product work. My journey from Android dev to AI was less a pivot and more an expansion of skills.
In enterprise sales, speed and relevance win deals. But when your best assets (case studies and specs) are buried in unstructured docs, the sales cycle grinds to a halt. I built a Sales AI to fix this.
Ad creation was a bottleneck. Dozens of formats, endless copy-paste, and little room for testing. I built an AI-driven generator that cut campaign creation by 98% and expanded testing capacity 5x.
People keep declaring prompt engineering “dead.” The truth is that prompts are to AI what compilers were to code: invisible, foundational, and the difference between toy demos and production systems.
Most spiritual education apps are cluttered with ads and require constant internet. I built Al Asma ul Husna as an ad-free, offline-first, multilingual app that organically reached #1 in its category with a 4.68-star rating.
Single clever prompts don’t scale. Enterprises need structured, reusable systems for reliability, governance, and iteration. Here’s how I design prompts as products, not hacks.
The DevOps department wasn’t broken. It was invisible, with no accountability, no metrics, and no trust. A forensic audit exposed waste and generated over $55,808 in measurable value.
Arbaeen attracts more than 20 million pilgrims annually. First-timers often struggle with basics like packing, navigation, and safety. I built Arbaeen GPT to provide multilingual, safe, and reliable guidance at scale.
Most AI translation fails with theology. It flattens nuance, loses reverence, and introduces errors. I built a Translation GPT tailored to Shia texts to preserve meaning, context, and sanctity.