Tatjana Chernenko
Applied AI & Data Scientist | AI Architect & Research Engineer | Enterprise AI Advisor and Leader
SAP SE, Walldorf / St. Leon-Rot, Germany
Academic background in Computational Linguistics and AI, Heidelberg University
Speech AI · ML · DL · RL · GenAI · NLP · Evaluation Architecture · Multilingual AI Systems · Enterprise AI · Agentic AI · Knowledge Graphs · RAG · Model Quality Optimisation
- 20 years of professional experience, including over a decade in AI and multilingual language technologies (hands-on, research and leadership)
- Applied AI & Data Scientist, SOTA Researcher, and AI Innovator specializing in Speech AI, NLP, LLMs, retrieval systems, knowledge graphs, and agentic AI.
- Hands-on research-to-production expertise in enterprise AI systems, combining advanced research, innovation, and state-of-the-art AI quality engineering.
- Leading complex enterprise AI initiatives from research to production at SAP.
- Deep pre-GenAI foundations in ML, Deep Learning, Reinforcement Learning, NLP, semantic clustering, neural generation, retrieval QA, summarisation, and task-oriented dialogue systems — extended into modern LLM, RAG, and agentic AI systems.
- Deep expertise in evaluation methodologies, model quality improvement, and domain adaptation.
Applied AI & Data Science & Research Engineering
Research-to-Production Systems · Evaluation Science · Enterprise AI
Evidence at a Glance
AI Innovation: Patents, Publications & Research Foundations
2016-present · SAP SE, Heidelberg University
Research-to-impact evidence spanning ML, DL, GenAI, NLP, Speech AI, multilingual systems, semantic retrieval, benchmark design, evaluation science, terminology intelligence, and production-grade AI quality improvement.
Enterprise Speech AI: Zero → Production
2024-2026 · SAP SE
Defined and built SAP's multilingual Speech AI capability from initial ambiguity to enterprise production across ASR, TTS, speech translation, voice workflows, and evaluation architecture.
TTS Evaluation & Pronunciation Quality Framework
2026 · SAP SE
Designed and implemented an automated TTS evaluation framework for enterprise speech generation, reducing dependency on large-scale human evaluation while making terminology-specific pronunciation quality measurable.
Terminology-Aware Enterprise AI Architecture: RAG, KGs & Agentic AI
2019-present · SAP SE
Shaped terminology-aware AI architecture and feasibility directions across RAG, knowledge graphs, agentic workflows, query reduction, and LLM-based/hybrid enterprise systems, grounded in SAP's large-scale multilingual terminology infrastructure.
OVKWS for ASR with Custom Vocabulary & Synthetic Benchmarks
2025 · SAP SE
Originated and led applied research with UNISINOS on customer-specific terminology recognition for enterprise ASR, combining open-vocabulary keyword spotting, synthetic audio, and hard-negative strategies.
ASR Evaluation & Quality Governance Framework
2024-2025 · SAP SE
Built a reusable evaluation and benchmarking architecture for enterprise ASR, connecting general transcription quality, domain-specific robustness, terminology recognition, and deployment-readiness logic across multilingual scenarios.
Multilingual Speech Data Product
2025 · SAP SE
Architected the data-centric foundation for SAP's multilingual Speech AI capability: a governed speech data product transforming raw enterprise audio into reusable AI-ready assets under enterprise constraints.
Japanese Enterprise TTS Optimisation without Extensive Fine-Tuning
2024 · SAP SE
Designed, co-implemented and validated a pipeline-based approach for improving Japanese TTS pronunciation and naturalness in terminology-heavy SAP contexts without extensive fine-tuning.
Semantic Retrieval & MLTR Evaluation Framework
2019-2024 · SAP SE
Designed and implemented the evaluation backbone for SAP's Multilingual Translation Repository (MLTR), operating across millions of verified translations, 40+ target languages, and 2,000+ language combinations.
Deep AI Foundations: ML, Deep Learning, NLP & Conversational AI
2014-2018 · Heidelberg University, Empolis Information Management
Built and evaluated pre-GenAI NLP and Deep Learning systems across semantic representation, unsupervised clustering, neural generation, extractive summarisation, and task-oriented dialogue management at Heidelberg University and Empolis Information Management.
What I Bring
I work where state-of-the-art AI research, evaluation science, and enterprise production reality meet. My strongest work is not only delivering AI systems, but shaping the research agenda behind them: identifying high-impact AI problems, formulating research questions, translating emerging methods into enterprise-relevant architectures, designing the evaluation science around them, and turning the resulting knowledge into reusable systems, patents, publications, and technical decision frameworks.
I combine hands-on applied research depth with technical leadership: defining AI quality standards, advising teams, leading ambiguous research-to-production initiatives, and connecting scientific ideas with measurable product impact. My focus is state-of-the-art AI quality: evaluation methodology, model quality improvement, domain adaptation, robustness, and the transition from advanced research to reliable enterprise AI. In parallel, I am preparing a PhD by prior publication, with a publication track focused on applied AI research, evaluation methodology, and production-grade AI quality improvement.
Selected Research-to-Production Work
Career Timeline
Lead AI & Applied Data Scientist / Speech AI
SAP SE, Walldorf/St.Leon-Rot, Germany
Built Speech AI capability (ASR, TTS, Speech Translation) from zero to production; improved models quality, adopted ASR and TTS for SAP domain; designed evaluation and quality-governance framework; architected speech data product; led OVKWS (open vocabulary) ASR research; contributed patents on STT fidelity and Japanese TTS.
- •Enterprise Speech AI from zero to production
- •Pending patents on STT fidelity and Japanese TTS
- •LREC-COLING 2026 accepted paper
Applied AI & Data Scientist / Multilingual Retrieval & Terminology Intelligence
SAP SE, Walldorf/St.Leon-Rot, Germany
Owned AI-facing multilingual retrieval and evaluation architecture; drove keyword-to-semantic retrieval transition; analysed 12M+ term pairs; contributed three granted US patents.
- •40+ languages / 2,000+ language combinations
- •36+ evaluation reports per cycle
- •3 granted US patents on terminology intelligence
Applied AI Developer / Data Scientist
SAP SE, Walldorf/St.Leon-Rot, Germany
Built a task-oriented conversational AI system for technical support from scratch: dialogue management, deep learning, and semantic search. Full end-to-end ownership.
- •End-to-end data pipeline for complex enterprise data (unstructured support tickets, HTML, XML), incl. advanced pre-processing, AI-based intent classification, filtering, and feature extraction
- •Architecture: Semantic retrieval on vectorized KB (domain fine-tuned Word2Vec embeddings enriched with one-hot encoded metadata)
Data Scientist / ML Engineer
Empolis Information Management, Kaiserslautern, Germany
Worked on NLP and information extraction for scientific paper search in the pharmaceutical domain.
- •NLP for pharmaceutical domain
- •Information extraction
- •Scientific paper search systems
Computational Linguist
Spiegel Institut
Worked on spoken-language processing for autonomous driving / vehicle communication scenarios.
- •Spoken-language processing
- •Autonomous driving communication
- •Speech and dialogue systems
BSc Computational Linguistics
Heidelberg University, Heidelberg, Germany
PhD track & accelerated MA track offered by supervising professor (Pr. Dr Riezler); declined due to SAP recruitment. AI-focused degree in Computational Linguistics with foundations in NLP, ML, deep learning, reinforcement learning, statistics, programming, formal semantics, syntax, and computer science. Academic research work: task-oriented dialogue systems, neural data-to-text generation, word sense induction, semantic clustering, and extractive summarization. Thesis focused on deep learning and retrieval-based Conversational AI for technical support.
- •PhD track & accelerated MA track offered; declined for SAP
- •NLP, ML, deep learning, RL, statistics, formal semantics
- •Research: dialogue systems, neural NLG, word sense induction, semantic clustering
- •Thesis: Deep learning & retrieval-based Conversational AI
CEO / Founding Partner
Quintessentially Estates (UK/UKR) & Avventura (Prague, Czech Republic)
Earlier entrepreneurial leadership; strategy, ownership, stakeholder management, and cross-functional execution.
- •Entrepreneurial leadership
- •Strategic business development
- •Cross-functional team management
Frontier Collaborations
Frontier Collaborations
International applied research collaborations across Speech AI, enterprise ASR, robotics, embodied AI, and voice-assistance platforms. Contributed Speech AI and applied AI expertise across multilingual speech interfaces, enterprise ASR, open-vocabulary keyword spotting, Indic-language ASR, voice-enabled robotic systems, domain terminology, model selection, evaluation methodology, deployment constraints, and production-oriented Speech AI for enterprise environments.
Partners: