Web & enterprise apps
End-to-end web platforms and SaaS: dashboards, multi-tenant, billing, and admin, built to scale.
Solution architect. From idea to production.
I build software, enterprise apps, and AI solutions that hold up in production, at real scale.
Open for new projects
From the first idea to production rollout, one engineering partner across the whole product.
End-to-end web platforms and SaaS: dashboards, multi-tenant, billing, and admin, built to scale.
Native and cross-platform apps with Flutter, Swift, and Kotlin, with a smooth native feel.
Distributed systems on Kubernetes: microservices, event streaming, and resilience at scale.
LLM and RAG features, agents, and automation built into real products, not just demos.
Connect your tools and remove manual work with n8n workflows, APIs, and webhooks.
Technical strategy, architecture review, and a clear roadmap from idea to launch.
The path every query takes through a production RAG system. Each stage is measurable, auditable, and tuned for latency.
Normalize, chunk, and clean source documents
Generate embeddings and index in the vector store
Hybrid semantic and keyword search
Re-order candidates with a cross-encoder
Grounded answer with source citations
Sub-50ms end-to-end, every stage observable.
Everything I ship sits on this foundation, chosen for longevity, not hype cycles.
Systems that survive millions of events per day. These are the areas I optimize for.
Products I have designed and built. Tap any card to open the live site.
hamkadr.ir
A marketplace connecting healthcare staff with hospitals and clinics for shifts and hiring, with profiles, filters, weekly scheduling, and a mobile app.
Visit sitemeezi.ir
An all-in-one SaaS for cafes and restaurants: QR ordering, POS, inventory, staff, and sales analytics, hosted in Iran.
Visit sitebargevasat.ir
A real-time multiplayer Hokm card game against people or AI bots, with leagues, rankings, daily rewards, and cross-device play.
Visit siteflatrender.ir
An AI studio that turns 1,200+ templates into platform-ready videos and images in minutes, billed by render-seconds. In beta.
Visit siteFindings from real engagements. Not translated articles, not hype lists.
Why BLEU and ROUGE fall short for RAG, and the metrics that actually drive decisions in real projects.
ReadHow to combine n8n with LangGraph to build auditable, debuggable autonomous workflows.
ReadThree anti-patterns I see in 80% of Vertex projects, and how we cut 60% of monthly spend.
ReadDeployment pattern with KEDA, GPU sharing, and request hedging for stable serving.
ReadUsing Gemini Nano and LiteRT for offline inference inside mobile apps.
ReadThe framework I build for CTOs, from use-case discovery to first production deployment.
ReadNo cost, no commitment. We map the use case, the constraints, and the next step together.