Proposal for the implementation of CellarAI at Hawke Winery Limited

Risk and Technology

Assessment 2

Appendices

Appendix A CellarAI Proposal

Subject: Proposal for the implementation of CellarAI at Hawke Winery Limited

We are pleased to present a proposal for a six-month trial of CellarAI, an innovative, sensory AI tool that integrates machine learning and Internet of Things (IoT) sensors to optimise the wine-aging and blending process. CellarAI is designed to help Hawke Winery Limited (HWL) create perfectly balanced wines with greater precision and efficiency, significantly improving the quality of HWLs wines and lowering operational costs.

In our view, CellarAI represents a game-changing opportunity for HWL to gain a competitive advantage in the wine industry, while avoiding the risk of being outperformed by competitors that use AI and fully harnessing the business potential of AI.

CellarAI is a state-of-the-art tool that mimics the human sensory experience of tasting wine. Using advanced machine learning and IoT sensors, CellarAI continuously monitors key chemical properties in aging wines, such as tannins, alcohol content, acidity, sweetness, and body. This data is used to predict the optimal aging timeline for each batch of wine and assist in creating blends that are perfectly balanced, ensuring the highest-quality product.

By analysing the aging characteristics and quality profile of wines in real time, CellarAI can recommend precise blending strategies that align with the desired flavour profile. In numerous trials, CellarAI has been shown to produce blends that meet or exceed the quality of blends crafted by expert winemakers in 96% of cases. This AI-driven quality can potentially transform HWLs production process. Its deployment has resulted in a shift in HWLs blend quality standards as it has eliminated human error in creating the right balance of acidity, tannins, alcohol, and fruitiness. This is a significant development, considering the increasing quality cost of unsaleable wine.

Key features of CellarAI:

• Machine learning algorithms: Its advanced algorithms learn from extensive historical data, adapting to various wine varietals and conditions over time to predict the ideal aging and blending strategies.


• IoT sensors: Its embedded sensors in the wine cellar track critical chemical parameters such as tannins, alcohol, acidity, fruitiness, sweetness, and body at each stage of the aging process.


• Real-time monitoring: It collects data continuously, allowing it to continuously analyse wine quality and provide winemakers with real-time insights to make data-driven decisions.


• Automated blending: It suggests optimal wine blending ratios based on real-time sensory data, leading to high- quality blends that requires less human intervention. This automation does not require any expert technological knowledge and is supported by CellarAIs 24/7 support service.


• Predictive ageing optimisation: It monitors the ageing process to determine the precise moment when each batch reaches its peak, optimising ageing times and reducing waste.

Benefits of CellarAI to Hawke Winery Limited

1. Reduced waste and higher quality wine production:

HWL currently faces significant wastage due to suboptimal blending and ageing processes, with many batches of its wine not meeting the quality standards necessary for sale under the HWL brand. These wines are then sold at a discount to supermarkets as cleanskin wine, which diminishes profit margins.

By implementing CellarAI, the winery will be able to create blends that are of consistently high quality, reducing the need to sell wine as cleanskin. CellarAIs precision in blending ensures that the final product meets or exceeds the quality of expert winemaker blends in most cases, resulting in higher-quality products that can be sold under the brand.

2. Faster blending process and reduced trial and error:

HWLs current process of wine blending heavily relies on human judgement, which is time-consuming and prone to error. With CellarAI, the blending process is streamlined by automating much of the trial and error involved. This will speed up production times, allowing HWL to achieve the desired blends faster and with fewer resources dedicated to the blending process. CellarAIs ability to analyse chemical properties and suggest optimised blends will reduce the need for repeated physical tastings, saving both time and effort.

3. Cost savings on labour:

The wine blending process currently requires significant labour input, especially from staff working in the cellar under Nimeshs supervision. This labour requirement is currently being filled with casual positions to allow for sufficient resourcing depending on the quantity of wine being produced in a given month. By using CellarAI, fewer manual interventions are needed, reducing the number of labourers and casual staff required by over 85%. This precision and automation of the tool not only saves time but also reduces labour costs associated with cellar operations, improving HWLs overall profitability.

4. Increased consistency and precision:

Traditional winemaking relies heavily on sensory evaluations by skilled winemakers, which can be subjective and inconsistent. CellarAI eliminates these variables by continuously monitoring objective chemical factors. The result is a consistently high-quality product that meets the standards expected by consumers, ensuring that HWLs reputation for excellence is maintained and enhanced, while also ensuring that customers will never be able to tell the wines were crafted by CellarAI.

Trial implementation plan:

Duration: 6 months

Objectives:

• Evaluate the effectiveness of CellarAI in optimising ageing times and blending


• Measure improvements in wine quality and reductions in


• Assess labour cost savings and process

Phases:

• Setup and calibration: IoT sensors will be installed in HWLs cellar. CellarAI will be integrated into existing systems, and it will undergo initial calibration with existing wine batches.


• Training and testing: CellarAI will be trained using historical data from HWL to calibrate the During this phase, CellarAI will help to monitor the wine ageing process and suggest blending ratios.


• Optimisation and adjustment: Based on trial outcomes, adjustments will be made to CellarAIs algorithms to further refine its ability to predict optimal ageing and blending ratios.


• Evaluation and reporting: At the end of the six-month trial, HWL will analyse the operational conditions before and after the use of CellarAI and use data-driven evidence to determine whether to permanently implement CellarAI.

We look forward to discussing the next steps and collaborating with you on this exciting initiative.

Sincerely, Tamati Jones

Appendix B Supply Chain Review

Summary:

Per HWLs request, we provide the below supply chain review for managements consideration. We were approached to investigate all suppliers that HWL has engaged in the past year.

All suppliers were given a questionnaire to complete. Suppliers that failed to return a questionnaire were contacted and interviewed to ensure their responses were captured.

The employees of each supplier were also given an anonymous questionnaire. A summary of the employee questionnaire responses is provided below.

General:

• This is the first supply chain review conducted on behalf of


• HWL sources glass bottles, corks and a limited number of grapes exclusively from overseas

Environmental Impact:

• 40% of HWLs suppliers advised they have appropriate waste management practices in place. These practices were heavily documented, and formal agreements were in place with local waste management providers.


• All of HWLs suppliers use environmentally safe packaging. This is one of HWLs key requirements when it considers new suppliers.

Social Impact:

• Staff employed for suppliers in countries at a higher risk of modern slavery advised there are no formal workplace health and safety policies in place.


• Staff in countries that are not at a higher risk of modern slavery advised there are extensive workplace health and safety practices with regular training given to staff.


• The local population of suppliers that provide grapes to HWL have advised that the vineyards of suppliers in countries at a higher risk of modern slavery are avoided due to the known usage of aggressive and extremely dangerous pesticides.

Governance Impact:

• Several suppliers have been involved in data breaches of their customers data (including HWLs data). The suppliers have advised that these breaches were minor and exaggerated. No action was taken by suppliers for any of the data breaches besides changing passwords for administrator accounts.