Developer interested in AI, machine learning, and building things that actually solve problems. I work across the full stack, but lately I've been spending more time on classification models, computer vision, and connecting all of that to interfaces people can actually use.
- ML / AI: classification models, computer vision, prediction with real-world datasets
- Frontend: Vue 3 · Nuxt · Vite · Vuetify · Tailwind · PrimeVue
- Backend / APIs: Laravel · Node.js · REST design · OAuth 2.0
- Infra / DevOps: Docker · GitHub Actions · Linux · HTTPS
- Data: visualizations · external API integrations · dashboards
| Project | What it does | Stack | Area |
|---|---|---|---|
| corn-leaf-disease | Disease classification system for small-scale maize farmers using computer vision | Python · Deep Learning · CNN | 🤖 ML / AI |
| fifa-world-cup-model | Predicts FIFA World Cup 2026 results using XGBoost, LightGBM and Monte Carlo simulation | Python · XGBoost · LightGBM | 🤖 ML / AI |
| dexcalidraw | Excalidraw with user accounts, authentication and real-time collaboration | TypeScript · Auth | 🎨 Full Stack |
| vynta | Desktop app for Windows to annotate, highlight and zoom into anything on screen | Rust · Vue 3 | 🖥️ Desktop |
| musycharts | Visualize your top genres and artists using the Spotify API | Vue · Spotify API · Charts | 📊 Data Viz |
| foundry-scan | Turns public discussions and trend signals into ranked micro-SaaS opportunities developers can evaluate and build | TypeScript · FastAPI · Playwright | 🔍 Tooling |
I like understanding how things work before building them. When something doesn't exist or I'm not happy with how it's done, I build it myself - applies to internal tooling just as much as projects with real trained models on real data.
Currently exploring more of the applied AI side: how to integrate models into workflows and products that are actually worth using.
| Portfolio | deras.dev |
| @daiv_09 | |
| dderas | |
| GitHub | github.com/daiv05 |