Medical Equipment Manufacturer

Building a forecasting model for a medical equipment manufacturer resulting in significant forecasting accuracy improvements allowing better procurement planning and contract management

Challenge

• Forecasting challenges when it comes to their supply chain
• Demand Planners working very manually with back and forth between planners, sales, and manufacturing
• Goal is to improve this with a more intelligent tool

Solution

• Built forecasting model with features included in Darts in combination with hierarchical reconciliation
• Solution would stream data directly from SAP

Impact

• Forecasting improvements of 10-60% compared with historical forecasting methods, depending on the time outlook in the model
• Helps procurement negotiate better contracts with suppliers and plan manufacturing accordingly

Internal Project

Combining Azure Cognitive Search & Azure OpenAI services to summarize key insights from documents and provide answers to medical-related questions

Challenge

• The recent release of OpenAI services on Azure has made a lot of noise in the broader world with interesting use cases being tested.
• Despite that, the industrialisation and implementation of such services in concrete business applications is still to see

Solution

• Articles relevant to a user question are identified and retrieved using Azure Cognitive Search.
• The retrieved articles are fed to Azure OpenAI services. The output consists of a summary with references to articles used in the answer.
• The scope of the retrieved documents and the chatbot responses is limited to the medical domain.

Business Impact

• Speeding up information retrieval and summarizing key insights from a large amount of structured and unstructured data
• On client demand, information can be extended to other public sources or using only internal sources of a company data

Multinational Consumer Electronics

Designed an LLM-powered workflow for the customer team to analyze user behaviour

Challenge

• Convert natural-language questions into SQL code, run it on the database, and return results to user
• Ensure accuracy and efficiency in processing complex queries while maintaining data privacy and security
• Provide training and support to the analytics team to effectively use the chatbot and interpret the results.

Solution

• Multi-step workflow breaking the problem into smaller tasks
• Dynamic selection of relevant examples shown to the LLM using a Vector database
• Fine-tuning the LLM on specific use cases
• Accuracy uplift through prompt engineering

Business Impact

• Democratizing data access & analytics to non-technical professionals
• Improved understanding of state-of-the-art approaches to LLM text-to-SQL task
• Improved ability to assess challenges and feasibility of reproducing the project for other text-to-SQL use cases
• Potential for increased efficiency and accuracy in generating SQL code from natural language queries

Advanced causal inference techniques revealed key drivers of user retention in a gaming platform’s feature ecosystem

Challenge

• A leading technology company’s data science team needed to precisely evaluate the causal impact of platform features on user retention
• The complexity of identifying which specific features positively or negatively influence user engagement required advanced analytical expertise

Solution

• Implemented advanced causal inference techniques to analyze feature impacts on retention rates
• Developed a comprehensive causal graph modeling approach
• Utilized machine learning causal inference algorithms, specifically X-learner methodology
• Conducted statistical analysis to isolate and quantify feature contributions

Business Impact

• Provided actionable, data-driven recommendations for feature development
• Enabled evidence-based decision-making for product enhancement
• Created a framework for understanding feature causality in user retention

Multinational Consumer Electronics Company

Adding an AI layer to the internal knowledge platform and automating answer generation for better information discovery

Challenge

• The customer has an internal platform where teams can share information in various formats, including blog posts, dashboards (e.g. Tableau), and experiment results (A/B testing results of proposed features)
• Time-consuming process of synthesizing information from multiple documents to answer specific questions
• One of the main technical challenges was the necessity to adopt and integrate the internal Language Learning Model (LLM) for the platform
• Managing and processing multiple files with ad-hoc formats to ensure consistency and compatibility across the platform

Solution

• Developed and implemented custom parsers to handle the diverse and ad-hoc file formats, ensuring accurate extraction and storage of information from diagrams and tables in dashboards
• Deployed a Retrieval-Augmented Generation (RAG) solution on top of the extracted data to generate quick answer drafts from the retrieved resources

Business Impact

• Created foundation for future AI-powered knowledge management solutions
• Decreased time spent in meetings asking for information that’s already documented
• Developed reusable components for RAG implementation in other internal tools
• Cross-referencing information from multiple sources automatically
• Reduced dependency on subject matter experts for basic information

Global Strategy Consultancy

Implemented a global dashboard using Palantir Foundry to unify views on project leads and streamline cross-organizational operations 

Challenge

• Missing a centralized, aggregated view on prospect research to aid employees in effective decision making in account lifecycles
• Work-around was extensive and complex communication efforts, which suffered from geographically dispersed employees 
• Deliver a solution that ensures effective collaboration for gathering valuable information and insight

Solution

• Rewired existing pipeline, for improved efficiency in speed, scalability, security and robustness.
• Updated data ontology to integrate new source systems
• Developed an efficient solution allowing the creation of a dashboard unifying views on project leads and statistics, requiring expertise

Business Impact

• Streamlined operations with an innovative dashboard and enhanced existing tools with new features, giving more power to the user. 
• Explaining, implementing, and enforcing best practices on platform usage 
• Introduced templates eliminating significant manual work, leading to increased efficiency and a decrease in errors

Commodity Trading Company

Developed a data-sharing platform on Foundry, enabling the client, its partners & customers to track & share their Scope 3 emissions across the value chain

Challenge

• Incoming EU regulations required client and their partners & customers to report on carbon emissions
• Deliver a scalable solution that can aggregate carbon emissions from multiple external companies, ensure data ownership continuation, handle big data volumes, accommodate future growth, protect sensitive data and adhere to EU emissions regulations

Solution

• Developed a suite of applications on Foundry: Data Input, Data Sharing, Benchmarking and Supply Chain Analytics
• The product is composed of a dedicated Foundry access for each participant

Business Impact

• Users are able to input Scope 3 emissions and ask others to share
• Dashboards enable granular emissions analysis along their supply chain 
• Platform is being monetized; onboarded their first paying user with potential for further growth

Multinational Consumer Electronics Company

Developing internal causal inference library consisting of 10+ algorithms and applied library to 2 use cases

Challenge

• Build causal inference library to be used by other data scientists in the future for various projects
• Enable activities such as churn analysis and A/B testing to be conducted

Solution

• Developed library in collaboration with internal stakeholders, did unite testing/documentation/user guide
• Worked on two related use cases needed for the analysis
• Unit8 implemented 60% of the algorithms for the project

Business impact

• Library enables data scientists to easily use multiple algorithms for causal inference through a unified API design
• 10+ algorithms implemented

US Law Firm

Building an application to streamline the due diligence process through a combination of automated web searches and web scraping

Challenge

• Client offers due diligence as a service
• Due diligence process is multi-tenant; users from multiple separate parties work together
• Process slowed down by back-and-forth communication and repetitive tasks
• Goal was to automate due diligence process

Solution

• Built a multi-tenant application to streamline the due diligence process end-to-end
• Implemented automated web search, website scraping and screening of entities linked to the target company

Business impact

• Project still in deployment, no impact available at this time

Swiss Raw Material Vendor

Demonstrated how a cloud platform can ingest different data sources and made dashboards and applications that could drive strategic decision-making

Challenge

• Client wanted to have a view of the current raw material inventory in the warehouse
• Need to forecast the inventory on customer side in order to optimize production and logistic cost and to increase customer satisfaction

Solution

• Data ingestion from different data sources including SAP, MeteoSwiss, and IoT devices
• Using Supply Chain offerings to define ontology
• Clean data
• Apply forecasting models through our Darts library

Business impact

• Showcase different forecasting models alternatives
• Optimized warehouse logistics
• Reduced excess inventory stock

World Wildlife Fund (WWF)

Developed a wildfire prediction system in collaboration with WWF to anticipate wildfire in Bolivian forests through a new fuel model based on geo data

Challenge

• Bolivian forests are at a high risk of wildfire
• Wildfire prediction system needed to anticipate how a fire might start and where it will spread

Solution

• Used publicly available data such as weather, climate, past fires, and geo data to model the wildfire spread starting at a certain point in time
• Created a new fuel model for Bolivia by transferring knowledge from available US data

Business impact

• A complete fuel model for the Bolivian territory is now available
• Applications created enable visualisation of past fires
• Discovered main pain points of ML with geo data and established a clear plan on how to tackle them and speed up future use cases

International Reseller

Building a tool to provide a 360° view over the whole supply chain, coupled with an alerting system when a certain material is reaching low stock levels

Challenge

• Recurrent struggle with materials being out of stock resulting in significant loss in sales

Solution

• Built application enabling users to see the whole supply chain, either globally or per material
• Implemented an alerting system that reports when a certain stock level is low and will soon need replenishment

Impact

• Reduction of average out-of stock levels at a given point in time by about 50%
• Lowered loss of turnover due to materials being out of stock

Nutrition startup

Onboarding a company on Foundry by providing personalised coaching and helping them implement an initial use case

Challenge

• Company recently acquired Foundry and the engineers lack the knowledge to launch an initial project

Solution

• Provided coaching to the engineers to teach them how to use Foundry’s capabilities
• Helped the engineers launch an initial project and create a minimum viable product

Impact

• Empowered internal engineers to use Foundry to its full potential on their own
  Completed end-to-end implementation of a minimum viable product, from data ingestion to dashboards

WWF (World Wide Fund for Nature)

Problem

• Wildfires can pose threats to ecosystems throughout the world, but in South America, the effects of deforestation can be particularly severe.
  organizations like WWF Bolivia and FAN constantly seek to improve their threat-modelling capabilities.

Solution

• Unit8 joined forces with WWF and Bolivian FAN NGO to leverage technology to help predict the spread of wildfires
  Our key objective was to identify and implement technological solutions that could supplement the existing work of WWF and FAN in the detection, modeling, and elimination of wildfire threats.

Impact

• The goal of the collaboration is to obtain faster reaction times to help authorities decide which zones and communities should be evacuated, and when.