Pinacle
Optimising food aid with AI-driven nutrition planning and privacy-preserving digital identity tools – turning surplus food into personalised wellness support for vulnerable communities.
Pinacle is a responsible digital solution designed to make food aid more efficient, transparent, and fair – connecting the logistics of food donation with the personal realities of nutrition and wellbeing. It demonstrates how intelligent technology and privacy-preserving data design can help food banks plan, distribute, and report in ways that are both operationally sound and socially responsible.
The project has developed an integrated, AI-driven platform that connects food banks, volunteers, and recipients through three interoperable tools: a mobile app for citizens, a web dashboard for food-bank operators, and a privacy-preserving backend that aligns donated food inventories with individual dietary profiles. Together, these components enable food-aid organisations to plan and distribute food more efficiently while offering recipients guidance that is healthy, culturally relevant, and respectful of their privacy.
The Pinacle project tackles the critical inefficiencies within Europe’s food life-cycle, from surplus donation to redistribution, where valuable food is often lost before reaching those who need it most. Conceived to reduce waste and strengthen nutritional equity, it introduces a smarter, data-driven matching process between available food stocks and recipients’ dietary requirements. Through its AI-powered planning and privacy-preserving identity framework, Pinacle brings traceability, trust, and personalisation to an area traditionally managed by manual coordination and goodwill alone.
Piloted with the Food Bank of Western Greece, Pinacle has shown measurable results: planning times reduced by more than 30 percent, improved accuracy in meal-to-inventory matching, and positive user feedback on transparency and usability. Built to European standards for cybersecurity, data fairness, and algorithmic accountability, it offers a replicable model for responsible digital transformation across the social food sector – turning surplus into sustenance and technology into trust.
WHY THIS MATTERS
Across Europe, food banks and charities have become one of the largest social infrastructures supporting citizens facing food hardship. According to the European Food Banks Federation (FEBA), its 351 member organisations and more than 43,000 partner charities redistributed over 830,000 tonnes of surplus food in 2024, reaching around 12 million people in need¹. Yet this represents only part of the picture. Many additional networks – including national alliances linked to the Global FoodBanking Network (GFN) – operate alongside or in partnership with FEBA members, suggesting that Europe’s true scale of food assistance is far greater.
Behind these numbers lies a deeper structural challenge. Inflation, geopolitical instability, and rising living costs are driving unprecedented demand for food aid, while supply chains and donations struggle to keep pace. Food banks must now serve more people with fewer resources, under increasing expectations for transparency, nutritional adequacy, and data protection. As FEBA notes, these organisations are “confronting a perfect storm: rising needs, dwindling in-kind donations, and the necessity of market purchases to bridge widening resource gaps.¹”
For recipients, this often translates into uncertainty, repetition, and meals that fail to match individual dietary needs or cultural preferences. For food-aid operators, it means lost time, under-used data, and difficulty demonstrating impact. The consequences extend beyond logistics: food insecurity and poor diet remain strongly linked to chronic illness, social exclusion, and loss of dignity among Europe’s most vulnerable groups.
Pinacle addresses this systemic gap by applying responsible digital technologies – artificial intelligence, self-sovereign identity, and zero-knowledge proofs – to make redistribution smarter, safer, and more equitable. By connecting donated food with verified recipient profiles, the platform helps operators reduce waste, enhance nutritional value, and ensure that personal data remains secure.
In doing so, Pinacle provides a practical model of digital responsibility within the social food sector – uniting efficiency with empathy, and demonstrating how technology can strengthen trust, inclusion, and wellbeing across the European food system.
- European Food Banks Federation (FEBA, 2024). Annual Report 2024 – Food Redistribuiotn in Europe. https://eurofoodbank.org/wp-content/uploads/2025/09/FEBA-Annual-Report-2024.pdf
While millions across Europe struggle to put food on the table, more than 59 million tonnes of safe and edible food goes to waste. FEBA tackles this paradox by working with our broad, trusted network of food banks to redistribute surplus food to those in need..
WHAT THE SOLUTION DOES
The Pinacle project turns the daily challenge of food redistribution into a connected, data-driven ecosystem that serves both citizens and food-aid operators.
Imagine a volunteer at a food bank in Patras preparing deliveries for the afternoon. Instead of relying on an impersonal manual process of allocation, they open the Pinacle dashboard, which shows up-to-date donations and stock data, mapped against nutritional analysis and expiry information where available. Next to each incoming product – rice, yoghurt, vegetables – the system suggests how it can best be allocated based on verified recipient profiles: families with young children, seniors with diabetes, or single adults needing balanced calorie intake.
For the food bank, this means decisions are guided by data rather than guesswork. The dashboard’s AI engine continuously analyses stock data, matching items with anonymised dietary information collected via the recipient mobile app. Volunteers see clear visual indicators and transparency labels that make it easy to understand how each recommendation balances freshness, nutritional value, and recipient needs, while the system ensures that every record remains compliant with privacy and consent rules.
At the other end of the chain, recipients use the Pinacle mobile app to register securely, review available food options, and receive personalised recipe suggestions and ingredient explanations based on what the food bank currently has in stock. These explainable outputs are limited to showing ingredient composition and dietary suitability, helping users understand why certain foods or meal combinations have been recommended. The app never reveals personal health data; instead, it works through self-sovereign digital identity credentials that verify eligibility and preferences without exposing sensitive details. Behind the scenes, zero-knowledge proofs validate each transaction, ensuring that data remains private while still allowing the system to confirm fairness and eligibility.
All information flows through a modular, service-oriented backend architecture that links inventory management, nutrition intelligence, and consent verification into one trusted workflow. The backend interacts with food-bank databases through interoperable APIs, exchanging real-time stock data with the AI-based Nutrition Recommendation module while applying privacy-by-design protocols for data minimisation and anonymisation. This design allows each food bank to integrate its existing ERP or warehouse systems, as demonstrated during the pilot with the Food Bank of Western Greece, where a live API connection synchronised donation records with the Pinacle platform. A distributed-ledger component adds resilience and traceability, providing a verifiable audit trail for key transactions such as consent and delivery, while keeping all personal data off-chain.
For operators, these integrations enable analytics and compliance-ready reporting, allowing teams to monitor efficiency, quantify food saved from waste, and assess how allocations align with basic nutritional targets. The result is a transparent data environment where operational decisions, regulatory reporting, and social-impact measurement coexist within a single, responsible digital framework.
Through its pilot deployment with the Food Bank of Western Greece, Pinacle demonstrated tangible operational and social benefits: allocation times cut by over 30 percent, fewer stock-matching errors, and clear gains in trust and usability among both volunteers and recipients. The process shifted from reactive to data-driven, helping participants see how technology can translate responsibility into daily practice. For recipients, the experience became more transparent and empowering – not merely receiving a box of food, but understanding how each item contributes to personal health and sustainability goals.
Beyond immediate impact, Pinacle illustrates how responsible AI and privacy-preserving identity systems can modernise the broader social-aid ecosystem. The same architecture can be replicated across municipal food-rescue initiatives and charity networks, or integrated with local redistribution schemes where fairness, efficiency, and dignity must coexist. In this way, Pinacle transforms food redistribution from a logistical necessity into a verifiable act of social trust and digital responsibility.
HOW YOU CAN USE IT
The Pinacle platform provides a modular, operationally validated framework that can be adopted, integrated, or extended by different actors in the social food-aid ecosystem. Each group can engage with the system in distinct but complementary ways.
For food banks and charity operators: Pinacle delivers a pilot-tested digital framework that food banks can adopt and scale to manage donations and recipient allocations more effectively. Its modular dashboard and backend services support day-to-day planning, stock monitoring, and traceable reporting, while remaining compatible with existing ERP or warehouse systems through interoperable APIs. Because the architecture is modular, food banks can adopt individual components – such as the allocation engine, consent management, or analytics dashboards – and expand integration over time as capacity grows. Designed with practitioners during the Western Greece pilot, the system enables step-by-step digitalisation that strengthens efficiency, transparency, and accountability without disrupting existing workflows.
For municipalities and social services: Local authorities can build on Pinacle’s architecture to coordinate food-aid partners across multiple sites or regions. Its consent and identity-management components make it possible to issue verifiable credentials for eligible households, ensuring fair access while protecting personal data. Integration with municipal databases or donor systems can automate reporting on surplus recovery, social impact, and food-waste reduction – aligning with EU policy frameworks on circular economy and social inclusion.
For developers and innovators: Pinacle’s modular backend and API layer provide a practical framework for integrating responsible AI and privacy-preserving digital identity into food-aid or donation platforms. Developers can adapt individual components – such as the nutrition-recommendation module, self-sovereign identity layer, or zero-knowledge proof validation service – within their own architectures. Technical documentation and architecture diagrams from the pilot are provided to support future reuse once repository licensing and anonymisation steps are complete (pending).
For research and policy organisations: Pinacle demonstrates how the Digital Responsibility Goals (DRGs) can guide digital innovation in social-aid systems. Its approach combines traceable transparency, privacy-preserving data management, and a clear methodology for measuring social and operational impact. Research institutions can use the platform as a case study in responsible AI deployment, while policymakers can reference its privacy-by-design, traceability, and accountability models when developing standards for digital welfare services or public food redistribution.
Together, these outputs make Pinacle more than a single technical system — it is a transferable model for making social-aid networks more efficient, transparent, and fair, while upholding dignity, privacy, and trust in every interaction.
DIGITAL RESPONSIBILITY IN PRACTICE
Pinacle has built privacy into every process that handles user data, consent, or verification. Each participant in the food-aid network interacts through a self-sovereign digital identity (SSI) that gives them decentralised control over what information they share and with whom. This SSI can hold two types of credentials:
- Verifiable credentials issued by authorised food-aid partners confirming eligibility.
- Self-declared attributes such as dietary preferences or health needs, stored locally and shared only with explicit consent.
Consent is obtained explicitly at registration and can be reviewed, modified, or withdrawn at any time through a dedicated “My Data” interface. This allows users to inspect what information is stored, renew consent, or request deletion, which triggers complete removal of personal data while retaining only anonymised records for statistical evaluation.
A distributed-ledger (DLT) layer records verifiable proofs of consent and eligibility, creating an auditable trail of key events without revealing personal information. All sensitive data remain off-chain, with only anonymised proofs written to the ledger via zero-knowledge proofs (ZKPs) verified by smart contracts. This ensures that verification is possible, but re-identification is not.
During the Western Greece pilot, the SSI + ZKP setup achieved a five-second average proving time for secure log-ins. By separating eligibility verification from personal or health data and enforcing strict data-minimisation, Pinacle turned regulatory compliance into an everyday user experience: privacy becomes a visible, interactive right rather than a background promise.
Data fairness in Pinacle is implemented as an explicit quality-control pipeline for both data and algorithms. The AI-based nutrition-recommendation module was trained and evaluated on representative datasets combining real food-bank inventory data with anonymised recipient profiles that spanned different ages, health conditions, and socioeconomic contexts. A three-tier bias-mitigation process was also applied to govern the data lifecycle:
- Data-bias assessment: Input datasets underwent regular quality audits to detect missing or skewed categories – such as over-representation of certain dietary types. Data-balancing procedures were applied where samples were limited, and nutritionists reviewed the resulting distributions for plausibility.
- Algorithmic-bias assessment: Formal fairness auditing of the model’s predictive accuracy and nutritional adequacy was benchmarked across user groups, maintaining less than 10 % variation in performance between demographic segments. Independent experts validated these tests using fairness metrics derived from the Assessment List for Trustworthy AI (ALTAI) framework.
- Human feedback loop: Within the mobile app, users and operators can flag meal plans that seem inappropriate or unbalanced. These reports feed into retraining and manual-review cycles, ensuring that fairness is continuously reviewed rather than fixed at deployment.
All AI outputs are subject to nutritionist validation before they appear in recipient or operator dashboards, and explainable-AI components visualise how each recommendation weights nutritional value, freshness, and stock availability. During the pilot in Western Greece, this framework reduced plan-stock mismatches and confirmed equitable treatment across recipient categories.
By coupling algorithmic monitoring with professional oversight and open feedback channels, Pinacle translates data-fairness principles into a repeatable development routine – ensuring that equitable food distribution is measured, auditable, and continually refined rather than assumed.
Pinacle approached algorithmic trust as a measurable development goal, combining explainability, ethical assurance, and continuous validation. Its AI-based Nutrition Recommendation Module applied a hybrid approach that merges rule-based nutrition logic with data-driven matching between food-bank inventories and anonymised recipient profiles. Each model iteration was benchmarked for consistency and reproducibility, ensuring that nutritional outcomes remained coherent even as stock data changed.
Development followed the Assessment List for Trustworthy AI (ALTAI) as a living checklist. Experts from seven disciplines — AI designers, data scientists, nutritionists, front-end staff, compliance officers, and management — participated in review cycles to detect anomalies, bias, or drift. Model performance and decision transparency were documented in internal audit sheets, forming part of Pinacle’s AI-governance framework.
Explainability was also embedded directly into the user interface through Explainable AI (XAI) components. Volunteers and operators can see visual indicators showing how recommendations balance freshness, nutritional value, and recipient needs. Nutritionists validated outputs before they appeared in dashboards, turning explainability into a practical oversight process rather than a background feature.
To maintain ethical integrity, Pinacle’s AI design adhered to the EU Ethics Guidelines for Trustworthy AI (2019) and anticipated requirements of the forthcoming EU AI Act. Algorithms were explicitly checked to avoid manipulative behaviour, coercion, or unintended influence on human autonomy. Feedback collected through the mobile and operator apps fed into retraining cycles, ensuring that interpretability and fairness improve over time.
By integrating human validation, cross-disciplinary review, and ALTAI-guided documentation, Pinacle demonstrates how explainability and accountability can be engineered into the core of an AI system — making trustworthy algorithms a verifiable practice rather than an abstract principle.
Transparency in Pinacle operates at both technical and organisational levels. A distributed-ledger (DLT) audit trail underpins the system, recording verifiable proofs of consent, eligibility, and allocation events. These logs are immutable yet anonymised, allowing donors, auditors, and municipal authorities to trace activities without accessing personal data.
Although governance is managed within a federated, permissioned network of participating food-aid organisations, data control remains decentralised. Each participant manages its own identity credentials and consent settings, ensuring that transparency does not compromise individual privacy or autonomy.
To reinforce system transparency, and in line with the project’s open-architecture design, technical documentation, architecture diagrams, and data-flow descriptions have been prepared to show how data move through the platform and how recommendations are generated. These materials are intended to support future external review once publicly released as part of Pinacle’s planned open-source publication.
At the governance level, analytics dashboards give operators and municipal partners real-time visibility of redistribution metrics, nutritional indicators, and compliance status. These dashboards meet EU-level reporting expectations for programmes such as ESF+, while strengthening accountability to funders and citizens alike.
For recipients, transparency is expressed more directly in the interface: the mobile app clearly distinguishes which items originate from the food bank and which come through partner retailers or other sources. Clear labelling and explanations in the app help users understand the composition of their aid package, plan household meals accordingly, and increase trust in the fairness of the distribution process.
Pinacle strengthens human agency by turning privacy controls into everyday choices. Building on its self-sovereign identity (SSI) and consent framework, recipients decide when and how to share information, can review what is stored, and withdraw their consent at any time. This design ensures that digital identity serves individual autonomy rather than system convenience.
Through these same mechanisms — explicit consent management, transparent audit trails, and clear feedback interfaces — users experience control as an active part of participation. Rather than data being handled on their behalf, they can remain aware of and involved in each exchange that affects them.
By linking privacy (DRG #3) and transparency (DRG #6) outcomes to user empowerment, Pinacle demonstrates how responsible identity and data systems can preserve dignity and agency within social-aid services.
CONTRIBUTION TO THE TOOLBOX
The Pinacle project contributes both technical and methodological assets, demonstrating how privacy-preserving digital identity and responsible AI can be applied within social food-aid systems. At this stage, the project’s public outputs focus primarily on documentation and methodological guidance, with open-source modules under preparation for release once software licensing and anonymisation processes are complete.
- Verifier Smart Contract for ZKP Verification: The verifier smart contracts developed within the PINACLE project enable the validation of ZKPs on-chain, ensuring privacy-preserving and secure access control to sensitive functionalities of the platform. These contracts allow decentralized verification of identity proofs without disclosing any personal data, thus supporting GDPR compliance and strengthening trust between food recipients, intermediaries, and food banks. By contributing this tool, PINACLE adds a reusable and scalable building block to the DRG4Food Toolbox for projects requiring privacy-preserving role-based access control.
View on DRG4FOOD GitHub
IMPACT AND OUTLOOK
The Pinacle project has demonstrated that digital responsibility and social inclusion can be combined in practice. Through its pilot with the Food Bank of Western Greece, the platform delivered tangible results: allocation efficiency improved by more than 30%, data-entry time was reduced, and recipients reported a more transparent and dignified experience. Volunteers and coordinators highlighted clearer communication, fewer mismatched supplies, and greater confidence in fairness across the process.
Beyond operational gains, Pinacle’s approach directly contributes to reducing food waste and advancing nutritional equity among vulnerable groups. By linking responsible AI, privacy-preserving identity, and transparent reporting, the project shows how technology can strengthen public trust in welfare and social-aid systems.
The architecture’s modular design enables reuse beyond the pilot region. Municipalities, NGOs, and other food-aid networks can adapt its framework to coordinate donations, manage inventory, and align redistribution with nutritional needs –without compromising privacy or dignity. The planned publication of its documentation, consent flows, and integration templates will further support this replication.
Looking ahead, the consortium will evaluate scalability scenarios and prepare a deployment roadmap with European food-aid partners. The project also aims to contribute to a European digital infrastructure for social-aid coordination and circular food systems, promoting efficient and equitable redistribution. In doing so, Pinacle provides a foundation for integration with broader EU frameworks on digital social inclusion, the circular economy, and responsible digital innovation – extending the DRG4FOOD vision into the social-impact domain.
QUICK FACTS
- Funding
- DRG4FOOD Open Call #2
- Use case
- Targeted Nutrition
- Partners
- Start date
- Oct 2024
- End date
- Sep 2025
- Resources