Knowledge Graphs · Semantics · Agentic AI
Semantics and reasoning are the missing layers in the current agentic evolution.
I write about ontology architectural designs, multi-modal data engineering, and building the neuro-symbolic operating systems that turn raw enterprise data into governed AI products.
Mission
My work sits at the intersection of ML, KG construction, and semantic modeling.
What drives me is the synergy between knowledge graphs and the fast-growing world of agentic systems: using the reasoning power of W3C standards to make LLMs more deterministic and trustworthy, and that integration far smoother to build. It starts earlier than most people expect, with the FAIRification of enterprise systems and the principles of Linked Data, because clean and meaningful representation is the foundation everything else stands on.
With those pieces in balance, my focus is on making this real in high-stakes domains like supply chain, automotive, industrial automation, construction, and utilities and energy. However, the exact same foundations hold anywhere decisions depend on data you can trust.
Trajectory
2023 - Present
Dresden, Germany
Own the semantic backbone of an AI document-processing platform. Built the ontology engineering practice and release tooling from scratch, made the case to build in-house rather than buy, and turned one-off customer models into shared domain assets that now steer the product roadmap.
2021 - 2023
Heidelberg, Germany
Where research met production. Built temporal knowledge graphs from live criminal, geospatial, and weather data for operational law-enforcement use, shipped the graph ML models that reasoned over them, and led NEC's ML work package in the EU Horizon project RAYUELA. Three patents and publications at AAAI, ECML, and NeurIPS came out of this arc.
2021 - 2022
Mannheim, Germany
Applied NLP and graphs at industrial scale: patent-extraction pipelines feeding enterprise search, and Neo4j property graph schemas wired into Azure ingestion pipelines with major query-performance gains on tens of millions of records.
2019 - 2021
Raubach, Germany
Built the data foundation for the controlling department, from ETL over messy PDF and OCR sources to APIs and dashboards used daily by 800+ sales representatives.
2019 - 2021
Koblenz, Germany
Graduate research spanning semantic web foundations, network theory, and ML security. Thesis work on generative adversarial networks laid the analytical groundwork for everything that followed.
2017 - 2019
Bangalore, India
First lessons in turning statistics into decisions: churn and financial-risk models delivered across corporate client engagements.
2012 - 2016
Bangalore, India
Foundations in embedded systems, signal processing, and logic design. Thesis work on intrusion detection in agriculture was an early lesson in applying technical rigor to unconventional domains.
Connect
I am always open to exchanging ideas, discussing new technologies, or exploring collaborative opportunities.