Summary
Sandra Castillo is a Senior Scientist based in Espoo with 17 years of experience applying deep learning and statistical methods to biological data, specializing in generative models for protein design and variational autoencoders. At VTT she has driven development of genome-scale metabolic model reconstruction and optimization algorithms, built open-source metabolic modelling tools (including AntND), and contributed to widely used LC-MS analysis frameworks like MZmine 2. Proficient in Python, Java, R and multiple compiled languages, she pairs hands-on ML engineering (TensorFlow/Keras) with backend and database work (Django, SQL, Neo4j). Her background in biochemistry and biology underpins a rare cross-disciplinary fluency that lets her translate high-dimensional omics problems into deployable algorithms. Beyond model building she has a track record of producing reusable open-source tools and practical web and data infrastructure for scientific workflows. Colleagues value her ability to bridge deep computational methods and wet-lab relevance, bringing research-grade models toward real-world metabolic and protein engineering applications.
17 years of coding experience
6 years of employment as a software developer
Master, Biochemistry, Master, Biochemistry at Universitat Autònoma de Barcelona
English, Spanish, Catalan, Finnish