<aside>
<img src="/icons/postcard_green.svg" alt="/icons/postcard_green.svg" width="40px" />
Professional Summary
Experienced Lead Data Scientist with a strong background in developing and implementing end-to-end machine learning solutions, MLOps practices, and driving innovation in the automotive and insurance industries. Proven track record in technical leadership, cross-functional collaboration, and delivering impactful data-driven solutions. Only data science relevant experiences are listed below.
</aside>
<aside>
<img src="notion://custom_emoji/de32dc2e-6882-41f4-b85d-f6aadb4914c2/12b329ca-38c1-8084-acca-007aba0016fa" alt="notion://custom_emoji/de32dc2e-6882-41f4-b85d-f6aadb4914c2/12b329ca-38c1-8084-acca-007aba0016fa" width="40px" />
March 2022 → present
- Led end-to-end development of EcoScore solution (scoring and coaching drivers on energy efficiency) for eco-fleet (fleet management for Paris Olympics), incorporating innovative feature engineering and MLOps practices.
- The solution led to 500k+ of total trips across 2.5k+ vehicles being scored on energy efficiency and each trip explained on what behavioural change would impact energy efficiency the most → ultimately helping achieve the goal of <40g/km emissions.
- Developed Multimodal ML solution to estimate EV Battery State of Health using contextual data including driving patterns, charging patterns, vehicle and weather conditions to provide actionable insights for a variety of uses cases, including prolonging battery life, residual value and fleet charing solutions.
- Spearheaded refactoring of Driving Analytics (an analytics codebase written in python that generates trip insights), transforming a big spaghetti codebase into modular and robust codebase, whilst also improving accuracy, efficiency, latency, and maintainability
- Led Driving Analytics project, making strategic decisions in product roadmap
- Established and implemented MLOps templates and practices covering entire model SDLC, including data & artifact version control, experiment tracking, deployment, model quality, data quality & drift, bias and explainability monitoring for customer-facing projects; whilst overcoming budget and resource constraints using open-sources stack.
- Led the initiative of Data Dictionary : an internal data catalogue enhancing company-wide understanding of the available data, and educating the importance of making data documentation across the company.
- Served as primary point of contact for several data science projects, handling development, L3 support, and improvements.
- Conducted several internal discovery, brainstorming and technical feasibility assessments identifying and developing POCs for potential business opportunities
</aside>
<aside>
<img src="notion://custom_emoji/de32dc2e-6882-41f4-b85d-f6aadb4914c2/12b329ca-38c1-801d-917b-007add16b87d" alt="notion://custom_emoji/de32dc2e-6882-41f4-b85d-f6aadb4914c2/12b329ca-38c1-801d-917b-007add16b87d" width="40px" />
Senior Data Scientist @ Kainos
October 2020 → March 2022
Kainos AI Practice (Internal IP Accelerators Development, and Workshops)
- Held several Azure Machine Learning workshops, with a wide audience
- Contributed towards developing internal IP accelerators for MLOps on Azure and AWS.
**FBD Insurance (Client Projects on Azure Machine Learning)**
- Involved in requirements gathering, led the discovery sessions to engage with key departmental heads to capture the processes and business challenges, and prioritise analytics opportunities, coaching and handover sessions.
- Developed and deployed Singularity Model, an AI based client matching solution to match and group identical quotes coming from the same customers, running as a service on Azure Cloud.
- Developed and deployed a Propensity Model to predict likelihood of conversion with high time-dependency features, along with the complete MLOps pipeline to automate regular retraining and re-deployment of the Model using Azure Machine Learning Pipelines and Azure DevOps.
- The project outcome was significantly increasing the CRM team’s efficiency by directing their focus and energy down from 300 quotes to top 5-10 customers to callback to on an hourly basis.
**Home Office (Client Project – Cloud Migration)**
- Worked closely with Data Engineers, Software Engineers and Technical Architects to update analytics codebase for migration from on-premises database & services to AWS Redshift.
**National Crime Agency (Client Project - Community Detection)**
- Leveraged Neo4j graph database and advanced graph analysis algorithms to develop a Community Detection solution, enhancing the agency's ability to identify and map complex criminal networks.
</aside>
<aside>
<img src="https://media.licdn.com/dms/image/D560BAQHY1EK0gg8_fQ/company-logo_200_200/0/1699377787807/kavida_ai_logo?e=2147483647&v=beta&t=fjlkYZhIObRXbgnBL7PDmDQHOqSQhu7riDsShgEt7SQ" alt="https://media.licdn.com/dms/image/D560BAQHY1EK0gg8_fQ/company-logo_200_200/0/1699377787807/kavida_ai_logo?e=2147483647&v=beta&t=fjlkYZhIObRXbgnBL7PDmDQHOqSQhu7riDsShgEt7SQ" width="40px" />
Senior Data Scientist @ Kavida.ai
March 2020 → August 2022 (Freelance)
- Led data science stream for AI-based supply chain resilience platform; helping build the internal data science capability
- Developed and deployed Bankruptcy Classification Engine to predict supplier bankruptcy
- Developed and deployed Threat Identification Model (Supply Chain) and profiling engine using NLP and AWS Lambda
- Established MLOps roadmap, templates and practices for continuous retraining and deployment on AWS using Sagemaker and MLFlow
</aside>
<aside>
<img src="https://companieslogo.com/img/orig/SNC.TO-5c1e9c05.png?t=1654067365" alt="https://companieslogo.com/img/orig/SNC.TO-5c1e9c05.png?t=1654067365" width="40px" />
September 2019 → October 2020
- Led predictive modelling of weather-based Weather Based Asset Failure Prediction Model for Network Rail, leading to predictive maintenance, minimising costs, delays and improving customer journeys
- Developed ensembled classification solution for asset failure prediction, deployed as a geo-spatial dashboard
</aside>
<aside>
<img src="/icons/user_blue.svg" alt="/icons/user_blue.svg" width="40px" />
Contact
Phone: 07970654781
Email: [email protected]
Address: London, UK
Web CV: bit.ly/raghav-dave
LinkedIn: https://www.linkedin.com/in/daveraghav/
</aside>
<aside>
<img src="/icons/graduate_pink.svg" alt="/icons/graduate_pink.svg" width="40px" />
MEng in Aerospace Engineering
Brunel University London, July 2016
</aside>
<aside>
<img src="/icons/archery_blue.svg" alt="/icons/archery_blue.svg" width="40px" />
Skills
- Cloud platforms: AWS SageMaker, Azure Machine Learning,
- Data Science: Supervised and Unsupervised Learning including clustering, classification, regression, time-series, graph / network analysis & data science, NLP, deep learning
- ML Frameworks: sklearn, xgboost, catboost, tensorflow, neo4j
- Tools: Power BI, Git, JIRA, Jenkins, Confluence, DVC, VSC, MLFlow, S3, Azure Blob
- Languages: Python, SQL, Bash
- Databases: Postgres, MongoDB, Athena, AWS Redshift
- Consulting: Presentations, Knowledge Transfer, Requirements Gathering, Consulting Sessions, Discovery Sessions.
- Leadership and Coaching: Managing a team of up to 5 to lead technical projects, mentor junior members
</aside>
<aside>
<img src="/icons/official-document_green.svg" alt="/icons/official-document_green.svg" width="40px" />
Certifications
- TensorFlow Developer Certification
- AWS Certified Machine Learning - Specialty
- Azure Data Scientist Associate
- Azure AI Engineer Associate
</aside>
<aside>
<img src="/icons/trophy_yellow.svg" alt="/icons/trophy_yellow.svg" width="40px" />
Key Achievements
- HackTrain VI Performance Challenge winner for developing tool to predict asset failures based on historic data.
- TDS Blogger
- Hosted Pie and AI Webinars
</aside>
<aside>
<img src="/icons/link_blue.svg" alt="/icons/link_blue.svg" width="40px" />
Learn More About Me
Projects
Tech Stack
Skills
Employers and Clients
Blogs
</aside>