Live Voiser
Built a browser-based event platform with QR access, live captions and translation in 125+ languages, sub-second delivery, transcripts, and AI summaries.
I build and productize web, mobile, and desktop experiences—from local open-source LLMs and production RAG to cloud operations, store launches, analytics, and growth.



Production products with real users, operational constraints, release cycles, and measurable outcomes. Most source code is proprietary; the live products are the proof.
Built a browser-based event platform with QR access, live captions and translation in 125+ languages, sub-second delivery, transcripts, and AI summaries.
Built a desktop AI workspace for macOS, Windows, and Linux with local LLMs, document RAG, clickable citations, editable artifacts, controlled tools, and air-gapped deployment.
Built an iOS and Android reading experience with EPUB, PDF, and TXT import, natural AI narration, summaries, personal libraries, and read-listen continuity.
Built the full short-form pipeline from topic research and sourced scripts to AI visuals, animation, narration, music, captions, and delivery for TikTok, Reels, and YouTube Shorts.
Built the iOS and Android product for legal drafting, file summarization, source-backed research, custom assistants, and UDF/PDF/DOCX export.
Built a desktop/on-prem product with locally running open-source LLMs, hardware-aware model distribution, offline deployment, and production RAG over 12M Turkish judicial decisions.
The common thread is ownership: choosing the right architecture, shipping the product, operating it, and learning from how people use it.
I carry one product vision into production across web, mobile, and desktop.
I build local inference, retrieval/reranking, and production RAG, with Turkish-legal fine-tuning experiments on compact 0.7–0.8B models.
I operate the services, data, and infrastructure layers behind AI products.
I do not stop at development; I release, measure, and build the growth loop.
I integrate coding agents across planning, implementation, debugging, testing, review, and documentation—not just prompting for snippets.
My scope expanded with the products: from shipping mobile apps to owning AI systems, cross-platform delivery, cloud operations, distribution, and growth.
Voiser AI
Directly building Live Voiser, Lokus, Clipa, and Voiser AI Book Reader while owning delivery across web, mobile, desktop, AI, and cloud layers.
Lawly
Directly building Lawly AI Mobile and the commercial on-prem product Lawly Local, including local open-source LLMs, model distribution, and production Turkish legal RAG.
OdaMobil Teknoloji
Directly developed 5+ Flutter products from idea and Figma design through monetization, analytics, store release, and growth.
I started in mobile development, then followed the products wherever they needed to go: web, desktop, cloud infrastructure, AI systems, app stores, analytics, and growth.
Today I am most useful where a product has to move from an ambiguous idea to a reliable system in users’ hands. I can reason about the model, build the interface, operate the cloud layer, ship to stores, and measure what happens next.
Clarify the user, constraint, business model, and success signal.
Choose the experience, architecture, AI approach, and operating model together.
Ship the right combination of web, mobile, desktop, backend, and cloud.
Handle stores, metadata, compliance, rollout, observability, and reliability.
Use analytics, feedback, and growth experiments to decide what comes next.
B.Sc. Computer Engineering · 2019–2023
GPA 3.3 / 4.0Led two engineering teams at Teknofest — Turkey's largest tech competition. Achieved 2nd and 4th place nationally. Managed multidisciplinary student teams through research, prototyping, testing, and reporting.
If you are building a web, mobile, desktop, or on-prem AI product and need someone who can own the whole delivery system, let’s talk.