--- title: "Laurynas Baltrušaitis (@laurynas26) | Hashnode" description: "Data insights in the wild. Read the latest articles by Laurynas Baltrušaitis on Hashnode." url: "https://hashnode.com/@laurynas26" --- @laurynas26 # Laurynas Baltrušaitis @laurynas26 · Amsterdam, Netherlands · Joined August 2025 Data insights in the wild ### [Building a House Price Predictor: What Held Up and What Didn't](https://laurynas-baltrusaitis.hashnode.dev/building-a-house-price-predictor-what-held-up-and-what-didn-t?utm_source=hashnode&utm_medium=feed) I started this project in September 2025 to answer a simple question: can I predict Amsterdam house prices? The build ran until the end of the year. This reflection comes six months later, delayed by 9 min read · [laurynas-baltrusaitis.hashnode.dev](https://laurynas-baltrusaitis.hashnode.dev/building-a-house-price-predictor-what-held-up-and-what-didn-t?utm_source=hashnode&utm_medium=feed) [Jul 28](https://hashnode.com/posts/building-a-house-price-predictor-what-held-up-and-what-didn-t/6a686efddbfbdb97fd4cae68) ### [Phase 5: From “Working” to “Correct”: Frontend-Driven Debugging and Finalizing Inference](https://laurynas-baltrusaitis.hashnode.dev/phase-5-from-working-to-correct-frontend-driven-debugging-and-finalizing-inference?utm_source=hashnode&utm_medium=feed) Up to the previous post, the project was in a reasonably good state: data scraping worked, feature engineering was stable, models were trained and selected via MLflow, and offline inference behaved as expected. The system seemed to do what it was sup... 6 min read · [laurynas-baltrusaitis.hashnode.dev](https://laurynas-baltrusaitis.hashnode.dev/phase-5-from-working-to-correct-frontend-driven-debugging-and-finalizing-inference?utm_source=hashnode&utm_medium=feed) [Jan 27](https://hashnode.com/posts/phase-5-from-working-to-correct-frontend-driven-debugging-and-finalizing-inference/6978dcf708c0679a6191a641) ### [Phase 4B Deploying the Price Predictor: What Worked, What Didn’t, and Why I Pivoted to a Hybrid Architecture](https://laurynas-baltrusaitis.hashnode.dev/phase-4b-deploying-the-price-predictor-what-worked-what-didnt-and-why-i-pivoted-to-a-hybrid-architecture?utm_source=hashnode&utm_medium=feed) Phase 4A was relatively smooth: I built a production-ready inference pipeline with modular preprocessing, caching, consistent schemas, and an MLflow-tracked model. Phase 4B was supposed to be the easy part: containerize the pipeline, wrap it in a Lam... 6 min read · [laurynas-baltrusaitis.hashnode.dev](https://laurynas-baltrusaitis.hashnode.dev/phase-4b-deploying-the-price-predictor-what-worked-what-didnt-and-why-i-pivoted-to-a-hybrid-architecture?utm_source=hashnode&utm_medium=feed) [Nov 23, 2025](https://hashnode.com/posts/phase-4b-deploying-the-price-predictor-what-worked-what-didnt-and-why-i-pivoted-to-a-hybrid-architecture/692340df5c30a3dc310e117f) ### [Phase 4A: Building a Real-Time, Production-Ready Real Estate Price Predictor](https://laurynas-baltrusaitis.hashnode.dev/phase-4a-building-a-real-time-production-ready-real-estate-price-predictor?utm_source=hashnode&utm_medium=feed) In real-world ML, an accurate model is only the starting point. The real challenge is delivering reliable predictions at scale, ensuring that every step at inference, preprocessing, encoding, and feature alignment mirrors training. In this phase, I b... 8 min read · [laurynas-baltrusaitis.hashnode.dev](https://laurynas-baltrusaitis.hashnode.dev/phase-4a-building-a-real-time-production-ready-real-estate-price-predictor?utm_source=hashnode&utm_medium=feed) [Nov 3, 2025](https://hashnode.com/posts/phase-4a-building-a-real-time-production-ready-real-estate-price-predictor/6908d06043c7fa07a021934c) ### [Phase 3B: Modeling Luxury Homes - Handling Outliers and Neighborhood Effects](https://laurynas-baltrusaitis.hashnode.dev/phase-3b-modeling-luxury-homes-handling-outliers-and-neighborhood-effects?utm_source=hashnode&utm_medium=feed) In real estate, extreme listings, both ultra-luxury and severely underpriced, can disproportionately influence predictions. These outliers, though rare, carry meaningful signals that must be preserved while keeping our model stable and robust. In thi... 11 min read · [laurynas-baltrusaitis.hashnode.dev](https://laurynas-baltrusaitis.hashnode.dev/phase-3b-modeling-luxury-homes-handling-outliers-and-neighborhood-effects?utm_source=hashnode&utm_medium=feed) [Oct 14, 2025](https://hashnode.com/posts/phase-3b-modeling-luxury-homes-handling-outliers-and-neighborhood-effects/68ee7b1f9f59b78bf99b3a61)