Ahtsham M.

Flutter engineer, five years in mobile, currently Senior Flutter & AI Developer at Naqvid. I take on the parts of an app that are hard to get right: architecture that survives a second developer, AI features that hold up under real usage, and backends that don't leak or fall over. What that looks like in practice: Architecture and state management chosen for the app's actual complexity, not for a blog post. Offline-first data flow, background sync, and cache invalidation that doesn't leave users staring at stale screens. Platform channels and native Kotlin/Swift when Dart runs out of road — camera pipelines, background services, secure storage, biometrics. AI that ships. LLM and RAG integrations where the work is retrieval quality, prompt cost and streaming latency, not the model choice. Voice pipelines (STT/TTS). On-device and edge ML when a round trip is too slow, too expensive, or too invasive — quantised models, warm-up handling, and honest fallbacks when inference fails. Most AI features die in the gap between a working demo and a feature users tolerate on a mid-range Android; that gap is where I do my best work. Firebase backends: Firestore models designed around the queries they'll actually serve, security rules written as the enforcement layer rather than an afterthought, Cloud Functions for anything a client can't be trusted with, and multi-provider auth. I also take on custom code and backend work inside FlutterFlow projects. How I work: I want the requirements pinned down before quoting, I break fixed-price scope into milestones you can verify independently, and I'll tell you when a feature isn't worth building or when a cheaper approach gets you the same result. Estimates are ranges with the assumptions stated. UTC+5, comfortable overlap with US and European hours. Recent work is under NDA — I can walk through the architecture and my role on a call.