Xebia

Transcript

1. Data strategy at Rainforest Alliance The conceptual information model as a means to understand, analyse and improve your data assets 11-06-2024 | Amsterdam | Ronald van Marissing & Steven Nooijen
2. Are you in control? 2 1. What data do I have? 2. What does it mean? 3. Where can I get it? 4. Can I trust this data? 5. Am I getting business value out of it? 6. Can I use this data? 7. Who (else) uses this data? 8. Is the data secure? 9. Is the data sensitive or confidential? 10. Am I in control? When you are able to answer these questions, holistically for your organization, you are on the right track:
3. 3 Who are we? Ronald van Marissing Rainforest Alliance Ronald is an experienced Technical Enterprise Architect with over two decades of experience in leading innovative technical solutions and driving operational excellence Steven Nooijen Xebia Steven is Head of Data Strategy and has an eye for organizational challenges surrounding the implementation of data & AI
4. 1. Company introductions 2. Rainforest Alliance’s digital ambitions 3. Rainforest Alliance’s information strategy 4. Our approach to creating an information model 5. Governance and next steps 4 Agenda
5. Company introductions
6. TYPE: ADVISORY Expert advice and guidance on how to successfully integrate Data and AI into your business DELIVERY: DATA PRODUCTS Tailored data products to discover hidden patterns, predict future events and optimize user experiences DATA PLATFORMS Robust and efficient basis to collect, store, and analyse your data, on- premise or in the cloud TRAININGS Comprehensive courses and learning journeys to upskill the workforce with the latest data and AI knowledge Driving Your Success with Data & AI Technology Partners • AWS Advanced Consulting Partner • Google Cloud Premier Partner • Microsoft Gold Partner • Databricks Premier Partner • DBT, SODA, Snowflake, Fivetran Pragmatic, tangible data strategies for a comprehensive approach to managing and utilizing data. State of the art data product implementations, e.g. stock optimization, churn prediction etc.. Ready to use data platform reference implementations Pragmatic training program, tailored to all levels from Executive to Data Scientist Product Expert Data Strategists, Data&AI Delivery Leads, Analytics Translators Expert Data Scientists Expert Data Engineers Expert Analytics Engineers Expert Data Engineers, Analytics Engineers and Platform Architects Expert Data & AI trainers People Facts • Data & AI frontrunner • 600+ expert consultants • Over 20.000 professionals trained • Open-source contributors • Community builders Results Migrated 1000s of workloads to the cloud and improved customer experience Executed global AI strategy and 20+ successful projects based on value-driven approach Developed a robust, petabyte-scale and cost-efficient Data Platform Built a cloud-based Data Platform to optimize flight-scheduling and reduce delay
7. Xebians love to write books about their passion: data We actively contribute to and are involved in the Nr. 1 Dutch podcast about data. Co-organizer and founder of several data events and data communities. We love to share knowledge with our customer on stage. Our mission To be an authority in the world of data & analytics We actively contribute to the open-source community because we believe it is the origin of much innovation in our field. We write white papers on relevant topics and are happy to share our knowledge and vision on certain topics through this channel. Contribute Write Share
9. THE CLIMATE EMERGENCY + SPIRALLING BIODIVERSITY LOSS + SYSTEMIC RURAL POVERTY + WIDESPREAD HUMAN RIGHTS ABUSES Solving connected crises that threaten our future Photo: Mohsin Kazmi 3
12. Why Digital (information) strategy
13. 6/11/24 14
14. Our strategy is rooted in our organizational vision and articulates our 2030 ambition to reach 100m producers through our Alliance. 2030 AMBITION STATEMENT* Our 2030 ambition is for 100m producers to join the Rainforest Alliance, transforming X% of global agriculture and forestry markets to [net positive impacts], and activating a global movement for XX investments in thriving landscapes by 2030. DRAFT OUR VISION A world where people and nature thrive in harmony. OUR MISSION To create a more sustainable world by using social and market forces to protect nature and improve the lives of farmers and forest communities.
15. To increase speed and scale of our impact, we will need to grow our alliance from 4 million to 100 million producers by 2030. • Why 100 million? 1. It will force us to truly innovate because incremental change is not enough to tackle urgent threats facing the world. 2. This is approximately 20% of around • 500 million producers working in • tropical forest areas. It’s • a tipping point • for system change.
16. Differentiated Advanced Foundational Set of Requirements Defined Data to Collect Optional Extra Data Bundles Off-the-shelf Projects to help progress up the Staircase Mechanism to Channel Investment Theme C Theme B Theme A To help farmers & producers to manage the supply chain we developed a Staircase model
17. Data Points Collected within our framework Theme Type Data Type Collection Mechanism Management General Legality and Record Keeping Binary – Capture Y/N Compliance On-Site Geo-location Boundaries, Production Areas, Protected Areas etc Remote Farming GMO Binary – Capture Y/N Compliance On-Site Soil Health Structure, Density, Moisture etc Combination depending on final requirement IPM Binary – Capture Y/N Compliance On-Site Social Freedom of Association Binary – Capture Y/N Compliance On-Site Environment Natural Ecosystems Binary – No conversion, no production in protected areas Remote Wildlife Binary – Capture Y/N Compliance On-Site Water Binary – Capture Y/N Compliance On-Site Waste Binary – Capture Y/N Compliance On-Site
18. Role & execution of (information) strategy
19. 20 Realizing RA 2030 ambitions through a Digital Ecosystem RA Digital Ecosystem will provide all needed building blocks to enable and accelerate teams in developing and integrating their services into the RA digital portfolio enhancing the customer’s experience. Rainforest Alliance Internal Systems Backend Systems Data Tech Platform Third-Party Service Provider Customer Facing Services IoT Web/Mobile App Stimulate Innovation Increase Revenue Extend Customer Reach Drones myRA Farmer Market Actor Supply Chain Actor Time-to-Market Time-to-Volume Our 2030 ambition is for 100m producers to join the Rainforest Alliance, transforming X% of global agriculture and forestry markets to regenerative practices and activating a global movement for XX investments in thriving landscapes by 2030 RA Team
20. A phased approach to realize a (Open) Digital Ecosystem Reaching a mature and usable platform to realize the 2030 strategy requires a dedicated transformation. Open Platform (e.g., incl. Web Components, APIs, SDKs, IAM) Third-Party Service Provider Digital Channels IoT Web/Mobile App Drones RA Marketplace Farmer Market Actor Supply Chain Actor RADE Internal Teams 1 2 3 2 3 4 Tech principles Digital Channels Foundational Building Blocks Broader Adoption 1. Define architectural guidelines and principles to be followed by delivery teams and assess current IT Architecture against them. 2. Define Digital Channels to reach target users and collect business valuable data. 3. Assess current initiatives to mine foundational Building Blocks (to be productized in the platform). 4. Define, Deliver, and Productize for broader adoption and external use. 21 Rainforest Alliance Internal Platform Backend Systems Data SaaS <<buy>> <<buy>> <<build>>
21. Digital Channels must include Data Collection Channels to collect data across the supply chain. While the short-term priority is to consistently digitize the envisionedinformation models, digitalization should be considered to providedifferentiating offerings to all users across the supply chain. Domestic Port Foreign Port Processing End User Freight Sea Freight Freight Freight Food Supply Chain Visibility Consumers are more interested in the origin and contents of their food. Trust & Transparency Real time tracking and monitoring of food supply chain increases trust between parties. Data-Driven Auditing Transitioning from traditional jurisdictional driven practices to data-driven approaches Landscapes On-site (aerial) sensing On-site (ground) sensing Farm Farm Remote Sensing 2. Current certification information digitized through on field surv eys can be digitalized through new on- site sensing technologies. 1. Strengthening the position of the current GIS services (Remote Sensing) in the organization can help quantify the nature impact and provide Objective and Key Results to innovation initiatives. Digitalization: transforming existing processes using digital technologies. Digitization: convert analog informat ion into digital form (e.g., through surveys). 3. Similarly to the Landscape and Farm monitoring, channels across the supply chain are env isioned to collect v aluable data through digitalization technologies. RADE Tech principles Digital Channels Foundational Building Blocks Broader Adoption 23
22. 24 Why invest in an information strategy? Our strategic ambitions require extensive use of information … … However, our current information landscape is scattered and immature … Data will enable us to: q Give 100 million producers insights on how they can improve their farming practices and how to get certified. q Support supply chain actors in doing result based claims as we trace in detail where goods come from. q Report to investors on how their investments and grants contribute towards their sustainability goals. q Create the myRA portal where we service our partners through a single platform across all our offerings. q Provide insights to internal stakeholders to optimize our offerings and achieve operational excellence. We face the following challenges: Our organization and hence much of our data is siloed and hard to combine across our offerings. We cannot share our operational data easily with other services or external parties (interoperability). There are no common identifiers and hierarchies across systems. We do not do Master Data Management. We have data quality problems and have no plan on how to improve it. There is limited ownership of data. There is a limited sense of urgency and a lack of understanding on how we structure and deal with data. … Hence, a practical vision and roadmap is needed to achieve our goals. What we are aiming for with this track: q How do we translate our strategic ambitions (incl. the farmer-centric staircase) into a practical roadmap? q What information needs do our stakeholders have (producers, SCAs, partners and internal employees)? q Establish the foundation for: o Data capture, entry and storage in operational systems o Improving data quality of key data objects (Master Data) o Integration and interoperability: sharing data internally & externally o Structuring, modelling and presenting data
23. Our approach The conceptual information model as a means to understand, analyse and improve your data landscape
24. 26 Understanding data flow and it’s touch points Operational systems Data modelling Insight creation Consuming applications 1. Data Information flow 2. Data interoperatbility flow We start by identifying business value and gradually work or way backwards to technical requirements. 1. Ideation – Collect and refine the need for information within the organization. 2. Information Model – Validate the existing information model: what key objects do we need data on? 3. Data Management - How will we collect, manage and store our operational data? 3 2 1 Scope order Integration layer Supply Chain Excellence Insights Delivery DAP Data Excellence Data Excellence DMERL Markets Innovation S&A Integration Partner Data Excellence
25. 27 Working towards an information model
26. 28 Information needs were gathered by identifying data use cases for user groups User groups Information needs Data collection Use cases 1. Focus on end users We followed a user-group-based approach to capture information needs. User groups were aligned with the myRA team. 4. Data collection pain points When reviewing the information objects in the information models with COs, POs and BAs, we asked how the respective data is gathered and if there are any known data (quality) issues 2. Use case workshops For each user group, use cases were gathered by relevant stakeholders. Scope For the domains in scope, all steps of the pyramid were covered in order to advise on information needs and data collection. For the business capabilities of Rainforest Alliance that were not included in the scope, there is still a ”blind” spot in the information needs that chould be added in the future 3. Information models Based on the gathered use cases, important information objects where defined and how they relate to each other
27. 29 Capture information needs through a simple user story format specifically for data use cases
28. Extract key information objects from the user stories and identify how this data is collected
29. Business people will be able to tell you where (quality) challenges in your data model occur Requirements The OutSystems Contextualized Framework (CFW) module is designed to (1) generate requirements and (2) gather requirements data using surveys. However, its slow performance may impact user satisfaction (goal #1) and make it unsuitable for collecting survey data (goal #2). Farming practices Similar data is collected from farmers on slightly different levels (group, farm, unit) leading to duplicate, conflicting and confusing information as well as an administrative burden for the certificate holder. Audit Much needed audit process and results information is captured through CAF, Excel and emails but is not timely available for RA. Further investigation is needed whether this is a process and/or tool challenge.
30. 32 Your information challenges translate into a concrete and tangible roadmap for better data Review your existing operation: A review of the conceptual information model and identification of the key information pain points to tackle. Then plan for the future: A prioritised roadmap with the information pain points translated into concrete actions including objectives, key results, timelines, stakeholders, and required budgets.
31. 33 Do you have an information model?
32. Governance and next steps Making sure the conceptual information model isn’t just a one-off PowerPoint exercise
33. 35 The benefits for different stakeholders Supply Chain Excellence (COs, POs, BAs) Data professionals Business Staff IT Architects Management 1 Understand the information landscape and interdependencies between data objects and sources ü ü ü ü ü 2 Align ownership: who is working on what and why? ü ü ü ü ü 3 Understand information pain points in data collection ü ü ü ü ü 4 New processes or system designs must comply with the conceptual information model ü ü 5 Logical data models and ready-to-use data marts in DAP must comply with the conceptual information model ü
34. 36 Making it interactively accessible as single source of truth through a tool like LeanIX
35. 37 Introduce an Architecture Board to enforce holistic evaluation of proposed solutions An Architecture Board should give a final GO on plans from each concept to make sure that proposed solutions fit in the bigger RA technology landscape. The architecture board should: • Challenge the business case of a solution in the light of the business process it supports • Review plans against the conceptual information model • Review how tools integrate in the to-be information landscape • Review the implementation plan to roll-out a solution across its users • Understand dependencies on internal team like the integration team or DAP From left to right: conceptual information model, integration layer, and data landscape to evaluate new solutions against
36. 38 Now: • MyRa build on Information model • UserGroups are core of valuestream and customer journey identification • Re-use, Buy, Build architecture principle made easier with understanding of Information model Future: • Further identification ownership of information by technical domains • Continuous refinement and enhancement of the Information model to support evolving data needs and strategic goals The present and the path forward
37. rainforest-alliance.org De Ruyterkade 6 1013 AA, Amsterdam, The Netherlands 125 Broad Street, 9th Floor New York, NY 10004 USA Thank you