Hi, I'm Kira 👋
I'm an NLP engineer, curious about how AI systems behave when things go wrong — when inputs are messy, users are unpredictable, or models are pushed outside of clean benchmarks.
I graduated in 2025 from the University of Lorraine in Nancy with a MSc in Natural Language Processing (CS Track). My background combines linguistics, machine learning, and software engineering, and most of my work today is around LLM-based systems, multilingual NLP, and RAG.
In my recent work, I've been involved in evaluating and testing large language models, looking at issues like hallucinations, bias, misuse, and robustness under perturbations. I helped build internal tools to test models before they're deployed.
I'm currently looking for opportunities in AI Engineering roles and excited to take on new challenges and expand the possibilities of what we can do with AI.
Experience
AI Engineer — opsci.ai
Sep 2024 — Present · Paris, France
- Designed and implemented end-to-end data pipelines for multilingual claim matching, ingesting and transforming large volumes of noisy social media text for downstream AI systems.
- Built and evaluated retrieval pipelines combining multilingual embeddings, reranking strategies, and instruction-tuned LLMs, selecting production-ready trade-offs between quality, latency, and cost.
- Integrated supervised contrastive fine-tuning of a multilingual embedding model (jina-embeddings-v3) to improve recall, strengthen positive/negative separation, and reduce language-driven bias on real-world data.
- Developed automated evaluation and validation pipelines for GenAI workflows (hallucination detection, red-teaming), used as quality and safety gates before production deployment.
Research Intern — CNRS (Multispeech)
Jul 2024 — Sep 2024 · Paris, France
- Extended an academic project on French prosody coaching by collecting and annotating speech data from native and non-native speakers using professional audio equipment.
- Evaluated pitch-tracking and TTS models (CREPE, FCN-F0, RMVPE, Tacotron 2, FastSpeech 2, Glow-TTS) on a custom speech corpus.
- Built data-processing pipelines for F0 contours and prosodic features (Praat-based), contributing to prosody-aware TTS and pronunciation-training prototypes.
Founder & Director — Start Smart
Mar 2020 — Jul 2024
- Founded and ran an online language school delivering English and Chinese courses to 50+ learners, managing curriculum design, marketing, and daily operations.
- Designed personalised training programs for individual and corporate clients, improving learning outcomes and user experience.
Education
MSc in Natural Language Processing
University of Lorraine · Nancy, France · 2023 — 2025
Engineering-focused master's program covering machine learning, NLP, language technologies, and production-oriented AI systems.
- Foundations of Natural Language Processing
- Machine Learning & Neural Networks
- Language Models and Text Generation
- Dialogue Systems, Chatbots & Question Answering
- Multilingual NLP and Low-Resource Languages
- Opinion Mining and Discourse Analysis
- Speech Recognition, Synthesis & Signal Processing
- Semantic Web, Ontologies & Lexical Resources
- Data Processing, Storage & Retrieval
- Software Engineering & System Design (UML)
Master's thesis: Robust Multilingual Claim Matching with Embedding-Based Retrieval and Contrastive Fine-Tuning (final grade: 16/20 · internship supervisor grade: 20/20)
View detailed coursework
- Neural Networks, Intelligent Systems, Reasoning in AI
- Symbolic AI and Knowledge Discovery
- Language Models and Generation
- Dialogue Systems and Conversational AI
- Speech Recognition, Robustness and Signal Processing
- Semantic Web and Ontology Engineering
- Data Engineering and Software Development
- Tutored Projects and Research Internship
BA & MA in Theoretical and Computational Linguistics
South Ural State University · Russia · 2009 — 2014
Comprehensive training in formal linguistics combined with early computational linguistics, programming, and cognitive science.
- Formal Linguistics: Syntax, Morphology, Semantics
- Computational Linguistics & Natural Language Processing
- Logic and Mathematical Foundations
- Psycholinguistics and Cognitive Science
- Programming Languages and Language Processing
- Translation Theory and Practice
Master's thesis: Comparative Analysis of Cantonese Dialect and Standard Mandarin (grade: Excellent)
View detailed coursework
- Phonetics, Phonology, Morphology, Syntax, Semantics
- Formal Models in Linguistics
- Natural Language Processing and Automated Language Analysis
- Psycholinguistics and Sociolinguistics
- Logic and Foundations of Human Knowledge
- Programming Languages
- Translation and Contrastive Linguistics
- Chinese and English Language Practice
Portfolio
Skills
Engineering & Data
- Python
- SQL
- Data ingestion and transformation pipelines
- REST APIs (FastAPI)
- Docker
GenAI & NLP
- Embeddings and vector search (FAISS, Chroma)
- Retrieval-Augmented Generation (RAG)
- Hugging Face ecosystem
- vLLM and GPU inference
- Multilingual NLP systems
Infrastructure
- AWS
- GCP
- Lambda Labs
🌍 Languages
- 🇫🇷 French — fluent
- 🇬🇧 English — fluent
- 🇨🇳 Chinese — advanced
- 🇷🇺 Russian — native
Contact
If you'd like to get in touch — feel free to reach out.