Open to opportunities

Simon Grah

@simongrah

Message

I build production AI and predictive systems that reduce churn, improve retention, and scale decision-making.

Ireland

Message

What I'm looking for

I'm looking to build and improve production AI, LLM, and predictive systems where business outcomes such as retention, expansion, operational efficiency, and trustworthy decision-making matter.

I've built production AI, machine learning, and predictive systems for LexisNexis, Chanel, TotalEnergies, Thales, and clients in energy, logistics, and B2B SaaS.

At LexisNexis, I designed and productionized evaluation pipelines for RAG agents, benchmarked GPT, Gemini, and Mistral models, and helped migrate classic RAG systems to agentic architectures with tool calling and hybrid search.

At Chanel, I deployed a secure internal GenAI chatbot serving around 1,000 employees and built RAG pipelines combining knowledge graphs and vector stores. My retention and expansion work has reduced monthly churn by 22%, increased detection 19-fold, and lifted upgrades 2.7-fold through prioritization.

Across more than 10 years in data modeling and seven years in software engineering, I've delivered ML lifecycle platforms, Spark pipelines, explainable AI research, and predictive maintenance systems. I bring business questions into testable signals and production-ready products.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

LexisNexis

Jan 2026 - Present (8 months)

Designed and productionized evaluation pipelines for RAG agents, including LLM-as-a-judge pre-filtering and multi-model benchmarking. Migrated RAG architecture to an agentic approach and ran validation loops with legal experts.

RAG LLM Agentic AI Evaluation Google ADK MCP Semantic Search BM25 LabelStudio

AI Software Engineer

Chanel

Jan 2024 - Dec 2025 (1 year 11 months)

Productionized and deployed a secure internal chatbot with guardrails serving ~1,000 employees. Built RAG pipelines combining knowledge graph and vector store, and coordinated cross-functional teams.

GenAI RAG Knowledge Graphs Vector Stores LLM Fine tuning NLP Azure AI Databricks

Data Scientist

CKDelta

Jan 2022 - Dec 2024 (2 years 11 months)

Developed end-to-end email processing systems with Azure OpenAI and Llama 2, predictive maintenance recommenders with Neo4j, and optimized container flows and truck scheduling on Databricks.

Azure OpenAI LLaMA 2 RAG Neo4j Predictive Maintenance Databricks Python SQL

MLOps Engineer & Data Scientist

TotalEnergies Digital Factory

Jan 2020 - Dec 2022 (2 years 11 months)

Owned the full ML lifecycle for industrial monitoring of FPSOs, set up CI/CD pipelines with MLflow, and delivered 9+ PoCs including predictive maintenance and supply chain optimization.

MLOps CI CD MLFlow Spark Databricks Predictive Maintenance Supply Chain Optimization Dynamic Pricing

Research Data Scientist

Thales

Jan 2017 - Dec 2020 (3 years 11 months)

Led technical development of a cyberattack detection platform processing 600 logs/s with Spark Streaming, designed graph-based anomaly detection algorithms, and researched explainable AI for European defense projects.

Spark Streaming Anomaly Detection graph algorithms Scala Python XAI Shapkit Cybersecurity