HostHive
A two-sided marketplace connecting companies with professional event and trade show hosts.
Read more about HostHive →AI engineer · researcher · creator
I build machine learning and LLM systems that run in production, not in notebooks.
What I am building now, as my own projects. Open any card for my role and the details.
A two-sided marketplace connecting companies with professional event and trade show hosts.
Read more about HostHive →Founder and product owner. I defined the marketplace, both sides of the experience (companies and hosts), and the roadmap, and I lead the build of the website and the mobile app.
A two-sided marketplace connecting companies with professional event and trade show hosts: a Next.js website and an Expo mobile app on Supabase.
A Next.js website and an Expo mobile app, both on Supabase.
In progress.
/work/hosthive
An AI travel companion for tourists, powered by routes that locals share in exchange for rewards at local cafés. Early stage.
Read more about Go →Founder. I am building Go in Berlin: the idea, the product and the AI companion.
Go is an AI travel companion that talks with tourists, learns what they like and plans real routes through the city, using routes that locals record and share. Locals earn points and redeem them for rewards at cafés. Gobie is the voice companion. Early stage, starting in Berlin.
Start small with a local route library: people record routes with live GPS, a person on the team reviews each one, and approved routes earn points. The talking AI companion for tourists comes next, built on those approved routes.
Early stage. The first version, a local route library, is in development in Berlin. No partner offers are live yet.
/work/go-gobie
A digital queue for cafés, restaurants and shops, built solo in 13 days for the Colosseum Crypto World's Fair 2026. No app download needed.
Read more about QSwapie →Founder. I built QSwapie solo in 13 days: the swap-your-place idea, the Solana payment verification and the web app. It was my first hackathon.
Guests scan a QR code at the door, take a place in line and can offer to swap places, with nothing to download. Built on Solana: non-custodial, devnet test. Aimed at Berlin venues such as cafés, restaurants and shops.
A Next.js web app that runs in the browser. The buyer's wallet signs one Solana transaction and the server verifies the payment on chain. Non-custodial: QSwapie never holds anyone's money. Tested on Solana devnet.
Web app live at qswapie.de. Non-custodial, devnet test (a test network): no mainnet, no live money.
/work/qswapie
Real-time sign language to spoken language, on the web and on the phone.
Read more about Hands Up →Founder. I started Hands Up to make conversation between Deaf signers and hearing people easier, and I lead its research, privacy-first design and the on-device recognition approach.
Real-time sign language to spoken language on the web and phone. Version 1 is ASL to English, with Arabic and Hebrew later. Private by design (on-device), built with the Deaf community.
On-device sign recognition, so conversations stay private, shaped together with Deaf signers.
In progress.
/work/hands-up
Roles and projects from before. Each card says what I did there.
Pic-Time
A real-time feature that turns a photographer's full shoot into a finished album.
Delivered a 10x improvement in inference throughput in automated album design · dataset curation across 50,000+ galleries
Read more about AI Album Designer →Designed and built the AI Album Designer as part of the AI team: the multi-model pipeline (face recognition and clustering, automated photo selection, saliency-aware smart cropping) and the inference work that lets albums generate while the photographer waits. Also shipped face-based gallery access and keyword- or face-based photo search. It started as an internship project and grew into a core AI team and product pillar.
Photographers upload a full shoot and get a finished album on demand. Alongside it: face-based gallery access and photo search on embedding similarity, served as REST APIs.
Multi-stage inference tuned for speed; near-duplicate rejection, album-wide consistency and keeping faces intact under aggressive crops. Photo search uses embedding-based vector similarity behind REST APIs.
Shipped and in production.
Delivered a 10x improvement in inference throughput in automated album design · dataset curation across 50,000+ galleries
/work/pic-time-album-designer
Morphysm · for Deutsche Oper Berlin
One of the first GPT-3 conversational products in Berlin's cultural sector, shipped as a Telegram bot.
27,000+ monthly users, achieving an 80% reduction in manual support volume
Read more about GPT-3 opera assistant →Built and deployed the assistant: designed the conversation flows, the prompt structure and the fallback handling, and built the React interface. Wrote the repository docs and video tutorials for onboarding.
A GPT-3 conversational assistant for Deutsche Oper Berlin, shipped as a Telegram bot that answered real audience questions.
Structured prompts and conversation flows, with fallback routines for the questions the model should not answer itself.
Shipped.
27,000+ monthly users, achieving an 80% reduction in manual support volume
/work/morphysm-deutsche-oper
aLifeForms
Experimental AI work for an artist: an online gallery of interactive, generative works.
Read more about Generative AI gallery →Co-built an online AI gallery of interactive, generative works that combine computer vision and audio, and ran face-decomposition experiments.
Interactive, generative artworks that combine computer vision and audio, for an artist's online gallery.
Real-time generative pipelines, packaged in containers for digital art installations.
Exhibited work.
/work/alifeforms-generative-art
NoCompany
An open-source web app that teaches networking and man-in-the-middle attacks visually.
Read more about Monster In The Middle →I built the app end to end, including Wi-Fi connectivity simulations.
An educational web app, supported by the ISC Project, that makes network attacks visible.
React and Material UI front end, Node.js back end.
Open source.
/work/nocompany-mitm
Gaza Sky Geeks
An application web app for Gaza Sky Geeks' coding academy, and an Arabic HR management app.
Read more about Academy and HR web apps →Built both apps end to end, from requirements to deployment.
An application web app for the GSG coding academy and an Arabic HR management app.
React, Node.js and Material UI.
Delivered.
/work/gaza-sky-geeks
EF English First
A platform that connected English students with teachers.
Read more about Online English-learning programme →I co-founded it and ran both technology (CTO) and operations (COO).
An online English-learning platform connecting students and teachers. The authoritative CV calls it the "Within 6 Months" programme at EF.
Details coming soon.
Details coming soon.
2018.
/work/ef-english-first
Independent research
End-to-end NLP research on Reddit data about how people perceive AI job risk.
Read more about AI job-risk and emotion study →Details coming soon.
Topic modelling, staged LLM-assisted labelling (Grok, then GPT-4o-mini, then Mistral) and fine-tuned classifiers for fear, anxiety and hope.
Topic modelling, staged LLM-assisted labelling (Grok, then GPT-4o-mini, then Mistral) and fine-tuned classifiers for fear, anxiety and hope.
Public repository.
/work/ai-job-risk-research
Finding kidney stones and their location in CT scans with deep learning and CLIP-based approaches.
A high-performance computing project.
Topic-aware autocomplete from an LDA topic model and a character-level LSTM.
First notes coming.
Build notes and research write-ups will appear here, dated, once they are real.