Research Engineer – Neuro-Symbolic AI (Knowledge Graphs) & Multimodal Assistant Systems
- Bengaluru, India
- On-site
- Entry Level
- Posted
About Bosch Group
Skills in graph embeddings, graph neural networks, Answer Set Programming/Prolog, SWRL or Drools, causal reasoning, agentic AI workflows, MLOps/cloud deployment, research publications, and patent drafting will be an added advantage.
Bosch Global Software Technologies Private Limited is a 100% owned subsidiary of Robert Bosch GmbH, one of the world's leading global supplier of technology and services, offering end-to-end Engineering, IT and Business Solutions. With over 27,000+ associates, it’s the largest software development center of Bosch, outside Germany, indicating that it is the Technology Powerhouse of Bosch in India with a global footprint and presence in the US, Europe and the Asia Pacific region.
Responsibilities
- Conduct cutting-edge research in Neuro-Symbolic AI, focusing on the integration of formal knowledge representations such as knowledge graphs with modern machine learning and deep learning techniques.
- Design multimodal data pipelines integrating text, structured data, and, where relevant, visual or sensor information.
- Design and develop knowledge-driven AI systems with natural language interaction capabilities, enabling explainable, trustworthy, and human-centered AI assistants.
- Integrate Neuro-Symbolic reasoning with large language models (LLMs) and multimodal foundation models using knowledge grounding, structured retrieval and reasoning-aware workflows to improve robustness, interpretability, and domain adaptability.
- Design, build, and evolve semantic assets — such as ontologies, knowledge graphs, semantic data models, and symbolic rules / constraints — that ground, validate, and explain neuro-symbolic AI systems in production-oriented research settings.
- Prototype, evaluate, and validate research concepts through experiments, benchmarks, and real-world use cases, translating research outcomes into scalable solutions.
- Collaborate closely with international, cross-disciplinary teams of researchers, engineers, and product stakeholders to apply research innovations to business-relevant scenarios such as product engineering, diagnostics, maintenance, and repair.
- Contribute to technology transfer, supporting the transition from research prototypes to production-ready systems in collaboration with software and product teams.
- Publish research findings in top-tier conferences and journals, file patents, and contribute to Bosch’s intellectual property portfolio.
- Stay up to date with the latest advancements in AI, machine learning, knowledge representation, and multimodal systems, ensuring Bosch remains at the forefront of innovation.
- Actively participate in internal and external research communities, workshops, and collaborations to foster knowledge exchange and thought leadership.
Requirements
- Ph.D. or M.S. or M. Tech from top Indian or foreign institutes (IITs, IIITs, IISc etc.) in Computer Science or a related field (e.g., NLP, linguistics, artificial intelligence, cognitive science)
- At least 3 years of relevant professional or applied research experience.
- Expertise in Neuro-Symbolic AI Architectures and Frameworks: Proven ability to design, implement, and integrate hybrid AI systems that combine machine learning with symbolic reasoning (preferably knowledge graphs) to address complex requirements.
- Named Entity Recognition (NER) and Relation Extraction. to understand user queries, extract insights from text, and contribute to the automatic construction and expansion of dynamic knowledge bases.
- Experience utilizing popular graph libraries (e.g., RDFlib, KGLab, and PyTorch Geometric) to develop and deploy graph-based ML algorithms, including link prediction, node classification, relation extraction and graph embeddings like Node2Vec.
- Proven ability to work collaboratively in cross-functional and international teams.
- Experience architecting or implementing hybrid reasoning and analytical engines capable of complex problem-solving, including causal reasoning, planning, explainable decision-making, and symbolic-neural inference.
- Architect and implement hybrid reasoning and analytical engines capable of complex problem-solving, including causal reasoning, planning, explainable decision-making, and symbolic–neural inference.
- Hands-on experience with enterprise knowledge graph and semantic web technologies, especially Neo4J, Stardog, STEADY including ontology modelling using OWL/RDFS, RDF data modelling, SPARQL query development, SHACL-based validation, reasoning/inference, and graph-based data integration.
- Working knowledge of symbolic query and reasoning approaches such as SPARQL, SHACL, description logics, logic programming, rule-based inference, or constraint-based reasoning.
- Ability to formulate research questions, design experiments, define evaluation criteria, run ablations, and interpret results with scientific rigor and strong reproducibility practices.
- Hands-on experience working with, prompting, fine-tuning, and evaluating Large Language Models (LLMs) and Vision-Language models.
- Strong understanding of Retrieval-Augmented Generation (RAG) paradigms.
Preferred
- Demonstrated proficiency in building robust data pipelines capable of integrating, cleaning, and preprocessing heterogeneous data sources
- Understanding of core NLP concepts and practical experience using state-of-the-art NLP frameworks, libraries (e.g., Hugging Face Transformers, spaCy, NLTK, Gensim etc.) for text processing, tokenization, semantic parsing, and language modeling.
- Proven experience in extracting structured knowledge from unstructured text using NLP techniques like
- Use visualization tools like Gephi, D3.js for Graph analytics and visualization.
- Strong understanding of knowledge representation and reasoning techniques, including ontology design, schema alignment, rule authoring, explainable inference workflows, and hybrid symbolic-neural evaluation methodologies.
