Open to opportunities

Minerva Singh

@minervasingh

Message

I build production AI, RAG, and machine learning systems for regulated, climate, and biodiversity data.

United Kingdom

Message

What I'm looking for

I'm looking to build reliable, production-grade AI and machine learning systems that solve complex real-world problems in regulated, public-sector, climate, biodiversity, and sustainability settings.

I've built AI workflows for Deutsche Bank, the Natural History Museum, MHRA, the Office for National Statistics, and DEFRA, turning complex documents and scientific data into usable, validated outputs. My work spans prompt-engineered LLM extraction, RAG, knowledge graphs, document intelligence, and production cloud deployment.

At the Natural History Museum, I designed a Neo4j biodiversity knowledge graph and graph-enhanced RAG workflows to extract and validate handwritten species records against repositories including GBIF. I presented this applied AI work for biodiversity digitisation and validation at the Living Data Conference in Colombia in 2025.

For MHRA and Deutsche Bank, I've developed secure, evaluation-led document AI pipelines using Amazon Bedrock, OpenAI models, Azure Document Intelligence, hybrid retrieval, hallucination detection, and human-in-the-loop validation. I focus on auditability, confidence scoring, semantic retrieval, and reliable extraction from regulatory, legal, and financial documents.

My earlier research and consulting work applies Python, geospatial machine learning, forecasting, computer vision, and cloud platforms to climate risk, energy, sustainability, public health, and humanitarian challenges. I've published 10 lead-author papers on AI, machine learning, and Earth observation for sustainability, and have supervised MSc and PhD students.

Experience

Work history, roles, and key accomplishments

Current

AI Consultant

Current

Deutsche Bank

Mar 2026 - Present (5 months)

Developed prompt-engineered LLM workflows for extracting structured regulatory information from complex financial and legal documents. Designed LLM-as-a-Judge evaluation frameworks and RAG-based pipelines for document understanding.

LLM Prompt Engineering RAG ROUGE L Levenshtein Distance Cloud Run

AI Consultant

Medicines and Healthcare products Regulatory Agency

Nov 2025 - Mar 2026 (4 months)

Architected a secure RAG pipeline using Amazon Bedrock for semantic querying of regulatory documents. Implemented hybrid search and evaluation frameworks for safety-critical outputs.

RAG TF IDF BM25 Azure Document Intelligence Precision@K

AI Engineer

Natural History Museum

Oct 2024 - Feb 2026 (1 year 4 months)

Designed and implemented a biodiversity knowledge graph in Neo4j and integrated it with LLM pipelines for graph-enhanced RAG. Developed agentic AI systems for automated verification and reconciliation of species records.

Neo4j Cypher LLM RAG OpenAI Anthropic Mistral Agentic AI GBIF

Data Science Consultant

Office for National Statistics

Oct 2024 - Oct 2025 (1 year)

Built ML pipelines on Azure ML Studio analyzing UK import-export trends. Designed anomaly detection models and Power BI dashboards for geopolitical disruptions.

Azure ML Anomaly Detection Power BI LLM Prompt Engineering

Energy Data Science Consultant

Hexapower

Mar 2024 - Sep 2024 (6 months)

Developed time-series forecasting models for UK energy consumption using SARIMA and GARCH. Implemented LSTM and XGBoost models to analyze energy demand drivers.

SARIMA GARCH XGBoost Time Series Modeling Python Data Visualization

AI & ML Engineer

Vertenetik

Nov 2023 - Sep 2024 (10 months)

Built custom CV segmentation models and deployed on GCP. Applied image ML to ecological applications.

Computer Vision Segmentation GCP Image ML

Postdoctoral Researcher

Nature-Based Solutions Initiative

Oct 2022 - Sep 2024 (1 year 11 months)

Built and evaluated Python-based geospatial hazard models for climate risk assessment. Developed a RAG model to extract data from corporate ESG reports.

Python RAG LLM Azure GCP Climate Risk

Sustainable Fuel Mix Data Science Consultant

Argus Media

Apr 2024 - Aug 2024 (4 months)

Engineered R-based time-series forecasting models for fuel consumption and energy demand. Developed automated ETL pipelines and interactive RShiny dashboards.

R Time Series Modeling ETL R Shiny Data Visualization

Geospatial Data Science Consultant

The Biodiversity Consultancy

Jan 2024 - Jun 2024 (5 months)

Built scalable Python/R pipelines for global biodiversity analysis and solar impact modeling.

Python R Geospatial analysis Biodiversity Solar Impact

ML Engineer

DEFRA

Aug 2023 - Dec 2023 (4 months)

Deployed MVP on AWS and Azure for recycling cost estimation. Integrated GitHub version control and PowerBI for dashboards.

AWS Azure GitHub Power BI Machine Learning

Climate Change Consultant

War Child Org

Jan 2023 - Nov 2023 (10 months)

Evaluated climate change impacts on internally displaced populations in Yemen using geospatial datasets. Developed an Arabic-language LLM workflow to synthesize qualitative data.

Geospatial analysis LLM Qualitative Data

Data Science Consultant

Imperial College London

Jul 2023 - Jul 2023 (0 months)

Contributed to an Africa-scale childhood nutrition program using geospatial machine learning. Applied ML and computer vision to analyze agricultural patterns.

Geospatial Computer Vision Machine Learning Policy Analysis