Celonis and AWS Drive AI Innovation at Global Ecosystem Hackathon

The Global Ecosystem Hackathon returned to Celonis for a fourth year in a row, showcasing how Process Intelligence (PI) is the critical foundation for enabling AI in the enterprise. Co-hosted by AWS in New York City, the event brought together Celonis partners and solution experts to drive innovation, creativity, and develop solutions for high-impact use cases. The hackathon was also hosted simultaneously in Munich, Madrid, and Bangalore where Celonis saw a total 259 participants across 84 teams, 40 unique partners, three regions, and incredibly, people joining from 21 different countries.

Throughout the day, partners explored how Celonis can be combined with AWS to build composable AI solutions that transform and continuously improve business operations. Each team worked hands-on with the Celonis PI Graph to create or refine a prototype app within their sandbox, C4C, or client environment. The focus was on real-world applicability–building something meaningful using their own data and use cases.

At the end of an intense day, teams had five minutes to pitch their solutions to judges from Celonis and AWS. Together, the judges picked three finalists, one from each region, where they will compete for the big win later this month. The global winner will be announced at this year’s Celosphere.

Here is a round-up of what the participants in New York City built:

EY - Incident Management

Opportunity: Proper preparation is essential for quick and effective incident response, be it a cyberattack, supply chain disruption, or corporate policy breach. Yet too often, preparation consists of manual, one-off drills that are so infrequent and siloed they fail to provide adequate training. Likewise, crucial information (e.g., response plans, contact lists, prepared messaging, etc.) are often buried in static playbooks, emails, and even IT tickets. This makes it difficult to ensure these documents are up-to-date, responders can access them quickly, and they have been reviewed and approved by the appropriate executive stakeholders.

Solution: The EY team designed a solution to improve incident readiness through three key categories: proactive readiness, automation, and board-level transparency. An AI-powered agent scenario generator uses real process data from Celonis to create continuous drill scenarios, from system outages to ransomware attacks. With AI Orchestration, these scenarios are automatically fed into an existing PI model, which triggers simulated workflows. This approach creates a closed-loop feedback system, ultimately enhancing board-level transparency.

OFI - Supplier Compliance

Opportunity: Manually tracking supplier compliance metrics can lead to higher costs, compliance risks, and a lack of real-time visibility into supplier performance. In response, many organizations are looking for a better approach—one that makes it easier to identify reliable partners and ensures that shipments are delivered on time.

Solution: The OFI team's solution uses AI and machine learning to analyze supplier performance and provide logistics managers with a comprehensive view of supplier compliance. The application has three main features. Supplier Analysis classifies 5,000 vendors into "gold," "silver," or "bronze" tiers to identify risks to on-time delivery. Order Analysis gives a detailed view of closed and active orders, using historical data to predict on-time delivery. Finally, a Real-Time Bot, integrated with third-party apps like Slack, retrieves information in real time and answers complex questions.

Read more: How a German pharmaceutical company is using Celonis to achieve full transparency over its order management processes.

IBM - Contract Leakage

Opportunity: Contract leakage, where the full value of a contract isn’t realized, can result in significant financial losses for businesses. A primary cause of contract leakage is the inconsistency between contract terms in documents (such as PDFs) and in a company's systems, making it difficult to get accurate data comparisons and insights.

Solution: This IBM team created an enhanced Annotation Builder to scrape data from PDFs, enabling real-time comparisons and insights to combat contract leakage. The solution uses a large language model to extract contract data from PDFs to identify discrepancies and compare it with information in the Celonis platform. Generative AI is also used to summarize comment histories for each ticket, a task that traditional annotation builders cannot handle due to the one-to-many relationship. The system then uses action flows and skills to download the correct PDF and convert it into text.

IBM - Organizational Transformation

Opportunity: Chief Information Officers (CIOs) frequently face immense pressure when tasked with overhauling their organizations, a complex and often years-long process. These transformations require CIOs to simultaneously lower costs, eliminate technical debt, and realize business value.

Solution: To help CIOs manage complex organizational transformations, a team designed an application that eliminates the unexpected and provides confidence. The tool offers executives an overall readiness score for their core business processes, helping them gauge their preparation for a major transformation. The application also uses AI to analyze process variants, providing meaningful summaries that help leaders understand what's happening and identify key considerations for the transformation.

Read more: How one of the world’s biggest food and beverage distributors are using Celonis to transform their business, process by process.

Winner: Bloomfilter - Generative AI Agents

Opportunity: Generative AI agents are often built with static, pre-configured instructions, which leads to an optimistic but unrealistic view of how they'll perform. As processes change and evolve, these agents aren't updated easily, making the implementation a complex, manual effort. While the results of their work are tracked, the "how", including rework, corrections, and human sentiment, is not. This creates blind spots that obscure the agent’s impact on both business processes and the humans who interact with them, leading to cancelled implementations.

Solution: Bloomfilter created the Universal Agent Optimizer to provide a continuous feedback loop that uses the Model Context Protocol (MCP) to feed real-time process data from the Celonis platform back to the agents. The application includes a dashboard to monitor key metrics such as rework percentage, automated work, and required human interaction. It also leverages AI to classify issues and help agents improve where users can receive feedback on execution quality and human impact.

Joining Bloomfilter are the other regional winners: OFI, who won in the APAC region with their Warehouse and Delivery Orchestration Suite, and Ventum Consulting, the EMEA winner, with their 'Room Zero' Emergency Department Optimization solution.

Discover more from AI Labs at Celonis

The solutions don’t stop here. At an AI Lab held in our New York City office, an American multinational investment bank won big for their use of our Prediction Builder, addressing know your customer (KYC) case resolutions.

Stay tuned for the 2026 AI Labs calendar, which will feature more customers and partners, more AI, and new innovations.