Bloomfilter launches app to synchronise AI & human workflows
Bloomfilter has announced the launch of the Celonis Agent Miner by Bloomfilter app, a new tool designed to help enterprises observe, govern, and optimise the collaboration between AI agents and human workers.
The company's app addresses an issue that many organisations face as they introduce AI agents into their workflows. Despite the increasing trend towards deploying AI helpers in knowledge work, a recent Massachusetts Institute of Technology (MIT) study has found that 95% of AI pilots do not progress to full production.
The main reason for this high rate of failure centres on the challenge of integrating AI-driven workers into day-to-day business processes and the lack of improvement in agent performance over time.
Bloomfilter's new app, developed in partnership with Celonis, is aimed at addressing these challenges.
Key features of the Celonis Agent Miner by Bloomfilter app include process-mining proprietary data held securely behind enterprise firewalls, integrating process data from both traditional system event logs and new agentic event logs, and feeding the resulting process intelligence back into the AI stack. This feedback loop is intended to enhance agent prompts through large language model (LLM) fine-tuning, improve vocabulary with Retrieval Augmented Generation (RAG), and refine decision-making with Model Context Protocol (MCP).
Erik Severinghaus, Co-Founder and Co-CEO of Bloomfilter, said:
"We are thrilled to partner with Celonis in using process intelligence to turn agentic confusion into orchestrated collaboration between humans and machines,"
Aaron Fulkerson, Chief Executive Officer of Opaque and a Bloomfilter customer, commented on the governance and security value of the new application:
"We understand the critical nature of governance and security to the AI transition. We are excited to leverage Bloomfilter's novel Agent Miner app to help govern our agents as we develop the Confidential Compute infrastructure necessary for trusted AI."
The Celonis Agent Miner by Bloomfilter app aims to solve coordination breakdowns in organisations attempting to merge AI agents with their existing teams. By providing a consolidated view of work through the integration of both human and digital activity logs, the app helps quantify the return on investment of AI-enabled processes, governs agents alongside humans in hybrid workflows, and applies process insights to improve the efficacy of agents over time.
Choon Aun Quek, Chief Growth Officer of KMS Technologies, also weighed in on the significance of the partnership and new solution:
"The future of software development, and business processes more broadly, lies at the intersection of the human and digital worker. Bloomfilter helps us be the best in the world at helping our customers transform and prepare for the agentic future."
The app will be made available as part of the Celonis Platform Apps Programme, which offers a range of pre-built, domain-specific solutions developed by partners with expertise in industry and process Intelligence.
Eugenio Cassiano, Senior Vice President of the Strategy and Innovation Group at Celonis, described the partnership and the role of process intelligence in the future of AI adoption:
"We believe AI won't achieve its enterprise potential without process intelligence. We're excited to see our partner Bloomfilter use this intelligence to pioneer a new frontier-using our platform to understand and govern the behavior of AI agents,"
Bloomfilter plans to use the Celonis Agent Miner app to demonstrate how process-aware agents can operate within various business functions, including the software development lifecycle (SDLC), with a focus on quantifying AI process ROI and effective governance of hybrid human-agent environments.
