Now in its third edition, the Dataiku Frontrunner Awards have returned to spotlight the extraordinary data and domain experts who are paving the way for Everyday AI.
The 2023 Awards received 84 submissions from Dataiku users across the globe. These show an impressive momentum for AI applications, at all scales and in all industries. Yet all participants share one common feature: All aim to build data into their daily operations, from advanced analytics to Generative AI.
Composed of Dataiku executives and industry experts, the jury carefully reviewed all submissions to elect the winners and finalists for the 2023 edition. Read on to discover their pioneering achievements!
The winners of the first award categories, related to democratization, have made it possible for more people to use AI in their organization and across their industry.
Best Data Democratization Program
As part of their learning university, Moderna had initially launched one AI literacy program, designed as a general introduction to AI and how its 850+ employees can leverage it in their role. Then, as employees started to identify a growing number of use cases, the AI Applied program was created to empower people without coding background to leverage AI/ML in a self-service manner to streamline or augment their day-to-day work.
- Finalist: Merck - A Holistic Approach for Enterprise-Level Data Democratization
- Finalist: PASHA Holding - Building a Data-Driven Culture Through Upskilling Employees and Community-Building
Scale of Impact
SLB decided to scale data science and AI in the organization by internally upskilling over 3,000 energy experts. A custom Dataiku portal was launched for SLB employees, offering modular learning, hands-on exercises, and certifications, with a gamification campaign to motivate employees to complete the courses. It took the HR team less than two years to reach 600+ certified data science practitioners and 5,200+ Dataiku academy certifications.
AI for Healthcare
In the healthcare industry specifically, Roche developed a predictive analytics capability to predict the need for Independent Data Monitoring Committees (IDMCs) in clinical studies. Now, the IDMC team can maintain and use the model independently, without relying on IT resources.
This approach promotes data citizenship and demonstrates the value of data science tools to non-technical users, enabling them to generate business data that can be used across teams. It also has the potential to expand beyond Roche’s internal scope, benefiting patient safety industry wide.
- Finalist: Moderna - Building AI to Generate Targeted and Actionable Medical Insights
- Special Distinction - Student Project: NYU Student - A New Perspective on the Status of Clinical Trials
AI for Finance
The mission of the Financial Industry Regulatory Authority (FINRA) is to ensure market integrity and protect investors. Analysts analyze petabytes of market data to efficiently detect any insider trading and unfair strategies. Dataiku has not only made data more accessible but also empowered non-coders to harness the power of large datasets that were previously out of their reach. They are now autonomous in analyzing data for their own needs, which greatly improves time-to-insight and enables FINRA to react faster as an organization.
- Finalist: Davivienda - Developing a Comprehensive Customer Score to Effectively Segment and Prioritize Clients
AI for Manufacturing
As one of the world’s top ten largest heavy equipment manufacturers, Doosan Corporation is leading the way to augment manufacturing operations with data. For instance, it has improved the accuracy of steel capacity prediction to 98%, by developing a dedicated model which helps their team make more informed decisions.
Using Dataiku as a low-/no-code solution, it has more broadly lowered the barrier to entry for data analysis skills, making it easier for the data analysis culture to spread within the group.
- Finalist: Braskem - Smart Gels: Classifying Gels in Polymers Through AI
- Finalist: KANEKA Corporation - Automated Drying Process Temperature Control With ML to Optimize Resources
- Special Distinction - B2B: ABS - Commercial Vessels Churn Prediction for Sustaining Long-Term Revenue
In the acceleration categories, the winners have used Dataiku to speed up the creation of AI capabilities, effectively augmenting or streamlining existing business processes. These are for a good cause, to increase ROI, or by using the latest techniques in Natural Language Processing (NLP) and Generative AI.
Best Acceleration Use Case
Novartis, a leading global medicines company, leverages Dataiku to streamline and automate its data pipelines for commercial analytics. This resulted in a significant reduction in data ingestion time, as well as in the manual efforts previously required for maintaining PowerPoint presentations and creating visualizations. Developers now allocate their time and expertise towards more strategic analytics for brand uptake, driving innovation and accelerating project timelines.
- Special Distinction - Best Use of Solutions: MandM - Using Customer Lifetime Value Scores to Understand Inherent Future Value and Deliver Personalized Experiences
- Special Distinction - Accelerating Insights: M1 - Building Company-Wide Data Literacy Through a Slack-First Approach
Best Positive Impact Use Case
Nonprofit One Acre Fund empowers smallholder farmers across East Africa through asset-based financing and comprehensive agriculture training services. By operating across different countries, the organization faced decentralized data management and relied heavily on manual processing via Google Sheets.
To address this, they established a data warehouse in Snowflake and integrated it with Dataiku, enabling real-time credit scoring, automated data processing, and enhanced data analysis. This transformation has resulted in increased operational efficiency, improved decision-making processes, and positive outcomes across various teams and countries — clearing valuable time and resources to maximize the positive impact of the organization.
- For-Profit Winner: Davivienda - Recommendation Machines: An Analytical Window to Financial Inclusion
DaviPlata, a cash management service launched by Davivienda, faced a challenge in offering financial services to its low-income users due to lack of information for creating classification models. They utilized Dataiku to run a field-aware factorization machine recommender system, combining various data sources to generate a dataset for modeling, which enabled them to offer products to these users.
The model increased the number of customers who could be offered a new product from 7.9 million to 12.1 million, providing financial services and offers to a low-income segment for the first time.
- Finalist: ALMA Observatory - Rising Data Science & Analytics to Meet Operational and Astronomical Demands With Limited Resources
- Finalist: MEWA - Forecasting Animal Diseases to Inform Prevention and Public Action
Best ROI Story
- Winner: Standard Chartered Bank - Data Informed Space Planning to Support Hybrid Working & Optimized Capital Allocation
In response to changing office attendance patterns and a focus on improving profitability, Standard Chartered Bank developed a Space Planning and Optimization Tool (SPOT) to streamline data integration and analysis. Dataiku played a crucial role in data sourcing and preparation, simplifying data availability, enabling faster data sharing, and facilitating better decision-making through statistical techniques.
SPOT has resulted in significant productivity savings, a shortened lead time for adapting office space, reduced property costs, and a shift towards data-informed decision-making for business teams.
- Finalist: SLB - Sizing Billion USD Well Construction Tenders Using Web Application & Machine Learning Models
Best Moonshot Use Case
Moderna faced the challenge of efficiently collecting and analyzing medical insights to combat infectious diseases like COVID-19. They partnered a pharmacist in the Medical Affairs team with data engineers to develop an AI model, which utilizes NLP in Dataiku to summarize medical insights and perform sentiment analysis.
This enables Moderna to respond quickly to changing sentiments, reducing analysis time from months to days and optimizing resource allocation. As a result, new insights are discovered faster and more efficiently, ultimately supporting the medical initiatives of healthcare providers.
- Finalist: SLB - Sizing Billion USD Well Construction Tenders Using Web Application & Machine Learning Models
- Finalist: bp T&S - ML-Driven Quantitative Trading
Best NLP Use Case
Norwegian insurance company Frende Forsikring has automated claims reporting, saving between 55 and 75 hours of work per month. The team trained a BERT model on 10,000 emails to predict the correct unit for incoming emails, achieving an impressive 98-99% accuracy rate to redirect claims to the right management unit.
An API, created with Dataiku's API designer, facilitates email forwarding, and a "human in the loop" approach continuously improves the model — which has become so efficient in just a few months, that the AI is now distributing all emails in the claim center. The process is already being replicated for automating other processes and helping Frende scale.
- Finalist: Aviva – Empowering Decision Making With ML: Sentiment Analysis and Topic Categorization for Customer Feedback
- Finalist: NEC Networks & System Integration Corporation - Zoom Meeting Transcription and Scoring Without Code
Best Partner Acceleration Use Case
- Winner: Standard Chartered Bank, Deloitte - Unlocking the Value of Data and Driving Business Outcomes
The Standard Chartered Digital Center of Excellence partnered with Dataiku and Deloitte to scale AI capabilities and maximize project delivery. Deloitte provided expertise and resources to help fill skill gaps, while Dataiku's platform enabled the team to quickly develop and deploy data-driven solutions.
The day-to-day change brought about by this collaboration includes a shift towards quality-focused agile development, improved collaboration between teams, and enhanced documentation of projects through a Dynamic System Delivery Methodology for data and AI.
- Finalist: BP, Wipro - Automating Competitor Insights for Real-Time Updates
- Finalist: i2e Consulting - An ML-Driven Process for Drug Order and Route Selection to Optimize Stock, Cost, and Resources
As part of the categories related to trust, the winners are developing safe and compliant AI capabilities for the enterprise.
Best Approach for Building Trust in AI
In the highly regulated finance industry, Macquarie Group partnered with Dataiku to extend the use of the platform to meet their business needs for autonomous, governed access to trade data.
The solution is a key enabler in meeting regulatory obligations, promoting confidence in their analysis processes, monitoring and reporting, and ensuring data is protected. Through trusted, accessible, governed data, the business is competitive in a fast-moving industry, can respond to evolving regulatory requirements quickly, and realize efficiencies in time and cost.
- Finalist: FSRA - AI-Powered Risk Assessment For Financial Services Regulation in Canada
- Finalist: OHRA - Empowering the Data Team With a Central Platform and Governance
- Special Distinction - Peta-Scale Data: FINRA - Implementing Self-Service Cloud Scalability Across the Organization
Best MLOps Use Case
This financial services institution implemented agile model deployment techniques, utilizing the Dataiku Govern Node for governance, and building customized conditional pipelines to support compliance. Dataiku played a crucial role in scaling their MLOps practice, enhancing model explainability and fostering collaboration among teams.
This resulted in increasing the firm’s overall efficiency rate for deploying compliance models into production by more than 90% and allowed Machine Learning Engineers to build end-to-end pipelines while writing up to 75% less production code by leveraging the functionalities of Dataiku.
- Finalist: ZS Associates - Efficiently Managing Hundreds of Models in Production
- Finalist: Detailresult - Fully Automating the Production of Daily Predictions for Over 100 Food Retail Stores
Last but not least, the Community Choice Award was attributed by Dataiku Community members to the submission they found the most inspiring. Drum rolls please to reveal the last winner…
- Winner: Aviva – Empowering Decision Making With ML: Sentiment Analysis and Topic Categorization for Customer Feedback
As a leading U.K. insurance provider, Aviva faced challenges in effectively analyzing vast customer feedback data to enhance their online customer experience. Manual analysis was time-consuming, error-prone, and lacked scalability. They partnered with Wipro to build a machine learning solution using Dataiku, which generated cost savings of approximately £10,000 per month, as well as time savings, including a reduction by 50% in the time to generate weekly reports.