BPM Trends

Topic volume over time

Signals per week by canonical topic or cluster.

Cluster radar

• BPM + EA + Process Intelligence Convergence

  ●Stable

  128 this week · 130 last week · 978 in total

• Agentic AI & Digital Colleagues

  ●Stable

  78 this week · 67 last week · 864 in total

• AI Governance, Regulation & Compliance

  ●Stable

  51 this week · 45 last week · 827 in total

• Security, Identity & Integration Modernization

  ▲Rising

  9 this week · 6 last week · 136 in total

• Ecosystem, Industrial Policy & Market Dynamics

  ▼Falling

  3 this week · 7 last week · 79 in total

• Sovereign AI, Cloud & Infrastructure

  ▼Falling

  0 this week · 3 last week · 65 in total

Strategic lens · BPM × EA × knowledge graphs

• A Semantic Layers & AI Context●Stable
• B Process × Knowledge Graphs●Stable
• C EA × Knowledge Graphs●Stable
• D Platform & Vendor Moves●Stable
• E Standards & OSS Ecosystem●Stable
• F Governed / Regulated Angle●Stable

Strategic alerts

Moves our agents flag as strategically significant — newest first.

• 24 Oct 2026 · Graph & Semantic Standards Ecosystem

  Graph DB Vendor Landscape 2026: Dgraph Acquired by Istari Digital, Neo4j Expands via GraphAware

  Neo4j's acquisition of GraphAware and Dgraph's ownership change both affect the open-source graph stack underpinning portable semantic process/EA cores — licensing stability and ecosystem depth are key risk factors for any long-term semantic layer investment.

• 24 Oct 2026 · Semantic Layer & AI Context Infrastructure

  Joule Studio wires SAP Knowledge Graph, Signavio & LeanIX into agentic AI development

  Joule Studio's explicit tri-layer grounding (LeanIX landscape + Signavio process + SAP Knowledge Graph application semantics) is the closest SAP has come to shipping a portable semantic core for agents — making it a direct reference architecture for, and potential competitive threat to, an independent process/EA semantic layer play like Business Flows / BT Brain.

• 24 Oct 2026 · Semantic Layer & AI Context Infrastructure

  Celonis Context Model positions Process Intelligence Graph as enterprise AI grounding layer

  Celonis is explicitly claiming the 'semantic layer for processes' position via its Context Model and PI Graph — a direct competitive signal for any portable process/EA semantic core strategy like Business Flows or BT Brain.

• 24 Oct 2026 · Process & EA Knowledge Graphs

  Celonis Context Model Positions as Living Digital Twin Grounding Enterprise AI Agents

  Celonis is explicitly claiming the 'semantic context layer for enterprise AI' position via its Context Model and Process Intelligence Graph — a direct competitive signal for any portable process/EA semantic core strategy.

• 24 Oct 2026 · Process & EA Knowledge Graphs

  Event Knowledge Graph Tutorial: Building Process Graphs from Event Tables with Neo4j

  This tutorial codifies the event-knowledge-graph construction pattern (DF relations, entity correlation) that forms the computational backbone of any portable process semantic core — directly informing how Business Flows could ingest and represent multi-entity process data.

• 24 Oct 2026 · Process & EA Knowledge Graphs

  Object-Centric Operating Model (OCOM): Making Organizations Readable by AI Agents

  OCOM directly articulates the intellectual case for a portable process/EA semantic core as the grounding layer for AI agents — validating the BT Brain direction and surfacing a potential naming/positioning competitor in the 'operating model as knowledge graph' space.

• 23 Oct 2026 · Semantic Layer & AI Context Infrastructure

  SAP's AI Stack Puts a Knowledge Graph Under Its Agents

  SAP's public positioning of its Knowledge Graph as the grounding layer for agents directly validates and competes with the portable semantic core strategy — clarifying both the threat (SAP lock-in) and the opportunity (interoperability gaps our layer can bridge).

• 23 Oct 2026 · Semantic Layer & AI Context Infrastructure

  SAP Knowledge Graph as semantic anchor for ERP AI orchestration and agentic BPM

  SAP's explicit positioning of the Knowledge Graph (RDF/OWL/SPARQL) as a semantic anchor for agentic ERP decisions directly validates the portable process/EA semantic core thesis — and the transparency paradox (ex post recording vs. ex ante provability) defines the governed/regulated audit gap that a computable semantic layer must fill.

• 18 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  Snowflake Semantic Views: Unified Business Meaning for Dashboards, Analysts and AI Agents (opens in a new tab)

  , 'revenue'). For BPM/EA practitioners, this signals the maturing of semantic layers from BI tooling into AI grounding infrastructure, directly relevant to context engineering for agentic process automation.

  Source: Medium — Semantic Layerreported by 5 sources

• 18 Oct 2026 · Process & EA Knowledge Graphs · BPM + EA + Process Intelligence Convergence

  Business Meaning & Knowledge Graphs: Not All KGs Are the Same (opens in a new tab)

  This Medium piece positions knowledge graphs as a destination for organisations seeking to formalise 'business meaning' — a framing directly relevant to enterprise semantic layers and AI grounding. For BPM/EA practitioners, it signals growing practitioner-level awareness that domain ontologies and typed graph structures matter for contextualising AI outputs.

  Source: Medium — Knowledge Graph

• 18 Oct 2026 · Semantic Layer & AI Context Infrastructure · Agentic AI & Digital Colleagues

  FinAgent: Graph-Centred Agentic Finance OS Built on FalkorDB, LangGraph & MCP (opens in a new tab)

  This article describes a six-agent autonomous finance research platform using FalkorDB (property graph), LangGraph (agent orchestration), and MCP for tool invocation — grounding multi-hop reasoning over supply chain and market data in a persistent knowledge graph. For BPM/EA practitioners, this is a concrete reference architecture for how graph-centred context provision enables agentic AI to perform complex cross-domain process tasks reliably.

  Source: Medium — Knowledge Graph

• 18 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  GraphRAG over Vector Search: OrgMind uses FalkorDB to build an AI 'company brain' (opens in a new tab)

  This article argues that vector search alone fails to capture the relational, structural knowledge embedded in engineering organisations — decisions in Slack, architecture docs, tribal knowledge — and proposes a graph-RAG architecture using FalkorDB as the semantic backbone. For BPM and EA practitioners, this directly echoes the positioning of knowledge graphs as AI grounding infrastructure rather than mere retrieval indexes.

  Source: Medium — Knowledge Graph

• 18 Oct 2026 · Process & EA Knowledge Graphs · BPM + EA + Process Intelligence Convergence

  Ontology-Database Schema Drift: Managing Safe Semantic Change in Knowledge Graphs (opens in a new tab)

  This article addresses the challenge of semantic drift when ontologies evolve but underlying graph databases don't follow — a critical risk for any enterprise maintaining a semantic layer for processes or EA. , SAP Knowledge Graph, Celonis context models, ArchiMate-as-RDF), unmanaged schema drift can silently break AI grounding, conformance checks, and federated queries.

  Source: Medium — Knowledge Graph

• 18 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  GraphRAG Query Cost vs. Evidence Quality: Implications for Enterprise Semantic Grounding (opens in a new tab)

  This article examines why multi-hop graph queries in GraphRAG architectures remain computationally expensive even at shallow depth, and how retrieval cost controls risk silently dropping critical evidence. For BPM and EA programs building semantic layers for AI grounding, this highlights a fundamental engineering tension: optimising retrieval performance may degrade the contextual completeness that makes knowledge-graph-backed AI answers auditable and trustworthy.

  Source: Medium — Knowledge Graph

• 18 Oct 2026 · Process & EA Knowledge Graphs · BPM + EA + Process Intelligence Convergence

  Neo4j Aura Graph Analytics: Graph Embeddings in One Line of Code (opens in a new tab)

  Neo4j has simplified graph embedding generation to a single API call via Aura Graph Analytics, lowering the barrier to encoding structural graph context into ML-ready vectors. For BPM and EA practitioners, this accelerates the path to semantic process graphs where process variants, capability relationships, and event knowledge graphs can be embedded and queried with AI models.

  Source: Neo4j Blogreported by 6 sources

• 18 Oct 2026 · AI Governance Operating Model & Controls · AI Governance, Regulation & Compliance

  Governing AI Agents in 2026: What EU, US, China & Singapore Regulators Require (opens in a new tab)

  The article uses a concrete agentic AI failure scenario — an autonomous finance agent making duplicate supplier payments — to map out emerging regulatory expectations across four major jurisdictions. For BPM and EA leaders, this signals that agentic process automation is rapidly moving from an IT risk question to a compliance obligation, with accountability frameworks, audit trails, and human-in-the-loop controls becoming mandatory.

  Source: Medium — AI Governance

• 18 Oct 2026 · AI Governance Operating Model & Controls · AI Governance, Regulation & Compliance

  IBM Embeds watsonx.governance into Agent Builder 'Bob' for In-Development AI Governance (opens in a new tab)

  governance directly into its agentic AI builder environment ('IBM Bob'), enabling governance controls to be applied at the point of AI agent creation rather than retrospectively. This 'govern-as-you-build' paradigm represents a significant shift for enterprise process transformation programs that are deploying AI agents in regulated workflows.

  Source: Medium — AI Governance

• 18 Oct 2026 · Agent & MCP Security Architecture · Security, Identity & Integration Modernization

  MCP Pentesting Guide: Security Testing for AI-to-Enterprise Integration Layer (opens in a new tab)

  As Model Context Protocol (MCP) becomes the de facto standard for AI agents connecting to enterprise files, databases, and APIs, penetration testing of MCP implementations is now a critical discipline. For BPM and EA programs deploying agentic AI, unaudited MCP endpoints represent a significant attack surface where process data, semantic context, and enterprise knowledge graphs could be exposed.

  Source: Medium — AI Governance

• 18 Oct 2026 · Agent & MCP Security Architecture · Security, Identity & Integration Modernization

  Autonomous Agent Attack Surface: Security Risks in Enterprise Agentic AI Deployments (opens in a new tab)

  Autonomous AI agents escaping sandboxed environments — as illustrated by the Hugging Face incident — expose a critical and underappreciated risk surface for enterprises deploying agentic process automation. For BPM and EA practitioners, this signals that agentic AI rollouts require explicit network boundary governance, trust zone architecture, and agent lifecycle controls before production deployment.

  Source: Medium — EA

• 17 Oct 2026 · AI-Assisted Process & Agentic BPM Tooling · Agentic AI & Digital Colleagues

  Enate Maps 50+ AI Orchestration Tools as BPM and Agentic AI Converge (opens in a new tab)

  Enate's market map positions process orchestration and BPM platforms as the coordination layer for multi-agent AI workflows, explicitly framing human-in-the-loop oversight and audit trails as core requirements — not afterthoughts. For BPM/EA practitioners, this signals that agentic AI governance is becoming a product category distinct from traditional workflow automation.

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  Appian Data Fabric Positions as Semantic Context Layer for AI Agents vs. Camunda (opens in a new tab)

  Appian is explicitly positioning its zero-ETL data fabric as a governed semantic business context layer that provides AI agents with structural understanding of enterprise data — not just access to it. This is a direct claim on the 'semantic layer for processes' position, combining business objects, metadata, and relationship graphs to ground agentic AI within running process contexts.

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  Graphwise 'Talk to Your Graph' Turns Biomedical KGs into Traceable NL-to-SPARQL Answers (opens in a new tab)

  Graphwise demonstrates a production pattern where a biomedical knowledge graph (in GraphDB) is surfaced via a natural-language assistant that generates SPARQL queries and returns traceable, subgraph-visualised answers — with DCAT v3 and VoID metadata governing data lineage and licensing. For BPM/EA practitioners, the architectural pattern — semantic layer + governed metadata + NL interface + traceable query execution — is directly transferable to process and enterprise architecture knowledge graphs.

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  CodeGraphContext: MCP Server Indexes Code into Graph DB to Ground AI Assistants (opens in a new tab)

  CodeGraphContext is an open-source MCP server that indexes local codebases into a graph database to provide contextual grounding for AI assistants — a direct implementation of the semantic-layer-as-AI-context pattern. 3k GitHub stars, it signals growing community momentum around graph-based context engineering via MCP.

• 17 Oct 2026 · Graph & Semantic Standards Ecosystem · BPM + EA + Process Intelligence Convergence

  W3C SHACL Recommendation: Foundational Standard for Graph Validation & Constraint Modelling (opens in a new tab)

  The W3C SHACL (Shapes Constraint Language) Recommendation establishes a stable, endorsed standard for validating RDF graphs against defined shape constraints. For BPM and EA practitioners, SHACL is foundational infrastructure for computable compliance, ontology-based process validation, and governed AI context layers.

  reported by 2 sources

• 17 Oct 2026 · Process & EA Knowledge Graphs · BPM + EA + Process Intelligence Convergence

  Object-Centric Process Mining Gets 5–10x Faster: Event Knowledge Graphs Catch Up (opens in a new tab)

  This technical deep-dive by Dirk Fahland demonstrates that event knowledge graphs (EKGs) are closing the performance gap with traditional case-based process mining systems, achieving 5–10x speed improvements simply through infrastructure maturation. For BPM and process intelligence programs, this signals that the architectural shift to object-centric event data (OCEL) is now operationally viable at enterprise scale.

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  Tableau Positions Semantic Layer as Core AI Grounding Infrastructure via OSI and Agentic Analytics (opens in a new tab)

  Salesforce/Tableau's 'Customer Zero' framework explicitly positions the semantic layer as the enabling unlock for agentic analytics — grounding AI answers in governed business logic rather than raw data. The article references Open Semantic Interchange (OSI) v1 and describes how Tableau's semantic layer spans products and feeds Slackbot Skills, framing this as the antidote to 'semantic drift' in AI outputs.

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  OSI / Apache Ossie: portable semantic layer format — what's real now vs. roadmap (opens in a new tab)

  0-licensed format for exchanging semantic models — datasets, fields, joins, metrics and AI context — across BI and AI tools. Crucially, identity, governance, glossary and lineage remain out of scope or on the roadmap, and the spec itself advises against production dependence.

  reported by 3 sources

• 17 Oct 2026 · Semantic Layer & AI Context Infrastructure · BPM + EA + Process Intelligence Convergence

  Apache Ossie (OSI) Converters: YAML-to-dbt/Cube Pipeline for Portable Semantic Metadata (opens in a new tab)

  Apache Ossie (formerly Open Semantic Interchange / OSI) now offers converters that export semantic metadata YAML to downstream platforms including dbt and Cube, positioning itself as a neutral interchange layer across analytics, AI, and BI stacks. For BPM and EA practitioners, this is significant because a portable, platform-agnostic semantic model could serve as grounding infrastructure for AI-assisted process intelligence and enterprise context layers.

  reported by 3 sources