Summary
Aleksey Tikhonov is a senior machine learning scientist and Head of Research in Berlin with 11 years of experience building generative AI, NLP, synthetic data, and graph-based systems. He leads research at Inworld AI on benchmarks, synthetic data generation, RLHF/DPO-style training, LLM evaluation, automatic prompting, and safety frameworks such as Constitutional AI, while publishing scientific work. Prior to this he combined product analytics and technical leadership at Yandex—designing AB-testing infrastructure, KPIs, and analytical tooling—and co-founded and led engineering at a data-security and fintech-focused startup. As an independent researcher he has collaborated with Max Planck, Utrecht, HSE, Yandex.Research and DeepPavlov, bridging academic rigor with product-facing research. He’s comfortable moving projects from prototype to production, from building DLP-style information-propagation systems and trading analytics to operationalizing LLM evaluation pipelines. A pragmatic polymath, he pairs deep algorithmic skills with hands-on product metrics and experimentation experience that often surfaces non-obvious user-behavior signals.
11 years of coding experience
11 years of employment as a software developer
Lomonosov Moscow State University
English, Russian, German