Gelareh Taghizadeh

Led end-to-end AI transformation, driving strategy, architecture, and delivery of enterprise-scale AI solutions across private equity, public, and private sector clients. Bridged technical leadership and client engagement to take AI from presales to production with measurable business impact.

• Built and led high-performing AI teams, establishing MLOps standards and scalable delivery practices to enhance project delivery
• Managed AI lifecycle, implementing AI Readiness Assessments and ROI frameworks to ensure alignment of initiatives with business value
• Delivered GenAI systems (LLMs, RAG, multi-agent, document intelligence) using production-grade MLOps (Kubernetes, Ray, Pinecone)
• Led the first GenAI system into production for Transport for London, contributing to a nationally recognised award-winning initiative
• Architected scalable, multi-cloud solutions across AWS (Bedrock) and Azure (OpenAI), integrated into enterprise platforms
• Embedded Responsible AI practices, including governance, evaluation frameworks, and explainability (SHAP, LIME) aligned with EU AI Act & GDPR

Developed deep learning and LLM-based career path modelling solutions to analyse trajectories for strategic workforce planning.

• Designed candidate matching systems using LLMs to align CVs with job requirements, improving matching accuracy and efficiency
• Implemented generative AI-driven search and ranking algorithms to improve candidate and job listing prioritisation.
• Applied advanced AI and generative AI to enhance talent acquisition and workforce planning processes.

Led and mentored high-performing team, aligning data science projects with product and business objectives to drive impactful outcomes.

• Designed and delivered end-to-end ML solutions, remaining hands-on across model development, debugging, and optimisation
• Built and led data science team, coordinating efforts of data scientists, data engineers, and ML engineers to deliver scalable AI-driven products.
• Built a personalised podcast recommendation system using fine-tuned transformers and deep learning, leveraging user behaviour signals
• Developed podcast summarisation models (T5, graph-based deep learning, topic modelling) that contributed to 10% increase in user retention.
• Implemented automated tagging using BERT, RoBERTa, and XLNet, improving content categorisation and discoverability
• Developed speaker identification capabilities using entity extraction, enabling speaker-level insights and recommendations

• Led multiple data science projects, overseeing experimentation, deployment, and optimisation processes
• Delivered end-to-end data science solutions on a content aggregation platform, applying NLP and machine learning to drive content understanding and recommendations
• Monitored and enhanced model performance, ensuring scalability, robustness, and continuous improvement
• Collaborated with cross-functional teams (product, engineering) to develop scalable AI solutions addressing business needs

• Developed machine learning and NLP models for large-scale text understanding, enabling personalised business research across sales, marketing, and policy use cases
• Built and optimised anomaly detection and classification algorithms, enhancing data quality and retrieval accuracy
• Collaborated closely with data scientists, engineers, and researchers to deliver scalable, production-ready AI solutions
• Improved web crawling and data processing pipelines to efficiently ingest and analyse large-scale web content
• Led R&D on intelligent web page segmentation, improving content extraction and structuring from unstructured web data

• Developed a specialised recommender system using deep learning and NLP to enhance company matching and discovery capabilities
• Applied supervised and unsupervised learning techniques to improve information retrieval within the discovery engine
• Executed keyword extraction, topic modelling (NMF, LDA), and NER to extract and present insights from unstructured data
• Pioneered AI-driven products, PoCs, and prototypes to enhance company discovery and ranking
• Partnered with engineering and product teams to transform research into scalable, production-ready solutions

• Achieved clustering performance on par with centralised methods, optimising balance between accuracy, efficiency, and privacy
• Designed privacy-preserving approaches that secured local data, facilitating valuable contributions to global insights
• Reduced communication overhead through efficient distributed algorithms, improving scalability across networked nodes
• Proposed collaborative clustering model for peer-to-peer networks, enabling distributed data sites to approximate global clustering while maintaining data privacy

• Delivered teaching across undergraduate courses in Artificial Intelligence, C++ Programming, and Data Structures, fostering student comprehension and enthusiasm for technical subjects
• Facilitated lectures and tutorials while providing student support to promote engagement and understanding of core technical concepts
• Supervised associate degree students, offering academic guidance and mentoring to enhance their learning experience and monitor academic progress

Delivered the first customer-facing GenAI product (“Pulse”) into production, leading cross-functional teams and defining a scalable multi-agent architecture

• Developed multi-agent architecture for TrendCurve AI (TCAI), enhancing forecasting capabilities and enabling on-demand insights at scale
• Led AI across a PE-backed, multi-product organisation, driving delivery in a fast-paced, outcome-focused environment. Managed a global team of 10+ data scientists and AI engineers, working closely with a wider engineering team.
• Built and scaled AI team, adding 6+ engineers and establishing a delivery-focused function
• Introduced AI-led culture and tooling (e.g., Cursor), embedding AI into engineering workflows and accelerating developer productivity across the organisation
• Established AI strategy and maturity assessments at portfolio level, aligning AI initiatives with business priorities across products