BARC Deal Radar – July 2026: Enterprise AI moves from experimentation to control

Databricks dominated July’s disclosed investment value with a round of ~$3bn (reported). The rest of the month was busier at a smaller scale. Control, security and compliance accounted for nine investments. Outside that group, Valarian raised capital for sovereign AI infrastructure and Bespoke Labs for agent training and reliability. BI and enterprise services produced the two acquisitions with disclosed prices.

This Radar examines 37 events closely tied to enterprise data, analytics and AI management. Another 30 events provide context from compute, model development and vertical AI.

July also exposed a transatlantic split. European companies generated more primary events, but US companies captured most disclosed investment value. Within Europe, the UK produced the larger governance and security rounds, while Germany combined smaller early-stage financings with the month’s largest disclosed European acquisition.

Compared with June

June’s Deal Radar covered 48 events and more than $13bn in fresh funding across the full tracked market. A single $12bn round drove most of that total.

Enterprise data and AI in numbers

The 37 primary events comprised 28 investments and nine acquisitions:

• Twenty-three investments disclosed values totalling approximately $3.53bn; five did not.
• Two acquisitions disclosed prices totalling $770m; seven did not disclose terms.

Databricks accounted for about 85 per cent of the disclosed investment value. Emergent followed with $130m, then Neo with $100m. Valarian and inforcer each raised $50m. These five deals made up approximately 94 per cent of the total, but the event count was spread across the enterprise stack.

Governance and security had the most rounds. Software development drew $185m across three disclosed investments. BI, data orchestration, performance management, enterprise knowledge and AI services saw smaller rounds and acquisitions.

Related markets

The other 30 events comprised 29 investments in compute infrastructure, model development and industry-specific AI, plus RealPage’s acquisition of Cherre. Twenty-eight investments disclosed values totalling approximately $7.16bn. One investment value and the Cherre acquisition price were undisclosed. Together, all 67 July events represented approximately $10.69bn in disclosed investment value.

A transatlantic split: European breadth, US capital

The regional split appears in both event count and disclosed value:

• Europe accounted for 23 of the 37 primary events, the United States for 12, and Canada and Israel for one each. The count uses company or acquisition-target headquarters from the tracker and assigns Perceptron’s dual UK/UAE listing to Europe.
• US companies captured approximately $3.32bn of the $3.53bn in disclosed primary investment value, compared with approximately $193m for European companies and $20m for Israel. Databricks contributed $3bn of the US total; without it, the gap narrowed to $320m in the United States versus $193m in Europe.
• Disclosed transaction prices were almost balanced: $400m for the agreed purchase of US-based Domo’s assets and $370m for the takeover offer for Germany-based All for One Group.

Europe did not present one uniform market. The UK produced Europe’s largest disclosed primary rounds: Valarian and inforcer raised $50m each, Risk Ledger $32.4m and Meticulous $15m. This was the region’s clearest cluster of later and larger financing, spanning sovereign AI infrastructure, managed security, supply-chain risk and software assurance.

Germany showed a different pattern. Its disclosed primary financings for ARC Intelligence, Finto, Tentris and Auxilius were all below $5m; its largest disclosed transaction was the $370m takeover offer for All for One Group rather than a growth round. France, Italy, Spain, the Netherlands and Belgium contributed smaller financings or strategic and acquisition activity, but no comparable national scale cluster.

July therefore showed European breadth through distinct national patterns, not a single European funding model.

Databricks raises the stakes in enterprise data and AI

Databricks sat apart from July’s other platform deals. The company confirmed a strategic round at a $188bn valuation; the $3bn amount was reported elsewhere and did not appear in its announcement. The investment covers a platform that spans data management, analytics, governance and AI development. That breadth puts Databricks in a different category from vendors focused mainly on models or inference.

Company	HQ Country	Value (US$)	Stage
Databricks	United States	~$3bn (reported)	Strategic investment

Agent governance, security and compliance

AI agents turn governance into an operational problem because they can initiate actions, not just produce output. Neo’s $100m launch focuses on inventory, attribution and runtime policy enforcement. The BARC Perspective on Neo places that approach at the application and infrastructure end of governance and argues that enterprises still need to connect it with top-down risk management and data governance. July’s deals address seven distinct control layers:

• Neo inventories agentic software, attributes actions to human and machine actors, and enforces policies at runtime.
• inforcer helps MSPs standardise Microsoft 365 security, compliance and Copilot management across customer environments.
• NAAIA translates AI regulations into inventories, risk classifications and governance workflows. Auxilius turns enterprise controls into continuously executable compliance processes.
• Risk Ledger tracks supplier and supply-chain exposure. Tangos automates financial-crime investigations and produces examiner-ready case files.
• Dawnguard embeds security and regulatory guardrails into cloud architecture design, infrastructure code and continuous validation.
• 8Layers combines identity posture management, threat detection and response, and continuous compliance across human, machine and AI-agent identities.
• Beelzebub deploys adaptive decoy services to detect, divert and analyse attackers.

These products address different parts of the control lifecycle. The practical challenge is integration: runtime policy must connect with managed posture, regulatory evidence, architecture assurance, identity protection, risk visibility and threat detection.

Company	HQ Country	Value (US$)	Stage
Neo	United States	$100m	Seed + Series A
inforcer	United Kingdom	$50m	Series C
Risk Ledger	United Kingdom	$32.4m	Series B
Tangos	Israel	$20m	Seed
NAAIA	France	$6.5m	Series A
Beelzebub	Italy	$3.42m	Seed
Dawnguard	Netherlands	$3.3m	Pre-seed
Auxilius	Germany	$1.4m	Pre-seed
8Layers	Spain	$1.1m	Pre-seed extension

BI, analytics and decision platforms

Progress made July’s largest BI move. It has agreed to pay $400m for substantially all of Domo’s assets, including its AI and data products platform. The deal would bring more than 2,400 customers and partnerships with cloud data warehouse vendors.

It also broadens Progress’s position beyond data tooling into analytics and AI products. Credible Data is building a semantic layer so analytics tools and AI agents use the same business definitions. Goods & Services bought Genware for its products and delivery expertise. Cherre stays in the table as a vertical comparison, although real estate sits outside BARC’s main scope.

Company	HQ Country	Value (US$)	Stage
Domo	United States	$400m	Asset acquisition agreed; closing pending
Credible Data	United States	$10m	Seed
Genware	United States	Undisclosed	Acquired by Goods & Services
Cherre	United States	Undisclosed	Acquired by RealPage

Secure data, AI and orchestration infrastructure

These seven deals concern the controlled infrastructure behind production AI. The investments cover four different layers:

• Valarian provides sovereign enclaves and control infrastructure for high-consequence AI workloads.
• DataBahn routes telemetry for security and observability.
• Perceptron is building a decentralised AI data network.
• Tentris works with connected data through a graph database.

The acquisitions follow different strategies:

• NetApp bought DataPelago to move its data platform closer to compute-intensive data and AI workloads.
• Infoblox has agreed to acquire Kentik, connecting network and DNS infrastructure with visibility into network performance and traffic.
• Prefect’s acquisition of Dagster brings two workflow orchestrators under one owner.

The first two acquisitions add adjacent infrastructure capabilities; Prefect consolidates within orchestration.

Company	HQ Country	Value (US$)	Stage
Valarian	United Kingdom	$50m	Series A
DataBahn	United States	$40m	Series B
Perceptron	United Kingdom / UAE	$6.5m	Strategic investment
Tentris	Germany	$2.2m	Pre-seed, including grant funding
DataPelago	United States	Undisclosed	Acquired by NetApp
Dagster Labs	United States	Undisclosed	Acquired by Prefect
Kentik	United States	Undisclosed	Acquisition by Infoblox agreed; closing pending

Finance and performance management consolidate

The finance deals focus on defined workflows and measurable outputs:

• ARC Intelligence and Finto are building new tools for finance teams. BrightAnalytics raised growth capital for reporting and performance management, while valuemize focuses on cost and value intelligence.
• French data and AI consultancy Keyrus acquired inlumi for its finance transformation expertise and European customer base. Vena has agreed to acquire Morpheo AI to add a governed context layer across planning, variance analysis and close.

Kelley Lynn Kassa examines Vena’s strategy and its execution risks in BARC Perspective: Vena Solutions To Acquire Morpheo AI.

Company	HQ Country	Value (US$)	Stage
ARC Intelligence	Germany	$4.3m	Seed
Finto	Germany	$3.4m	Seed
inlumi	Netherlands	Undisclosed	Acquired by Keyrus
Morpheo AI	Canada	Undisclosed	Acquisition by Vena Solutions agreed; closing pending
BrightAnalytics	Belgium	Undisclosed	Majority growth investment
valuemize	Germany	Undisclosed	Pre-seed

Software development and assurance

July’s software deals cluster around two needs:

• Emergent turns prompts into production applications, while Anaconda’s acquisition of Kilo Code moves the Python data and AI company into coding agents.
• Bespoke Labs builds company-scale reinforcement-learning environments to train and evaluate long-horizon agents before deployment. Meticulous automates testing for software that teams can generate faster than they can review it.

The first pair targets development speed; the second addresses reliability and assurance.

Company	HQ Country	Value (US$)	Stage
Emergent	United States	$130m	Series C
Bespoke Labs	United States	$40m	Seed + Series A
Meticulous	United Kingdom	$15m	Series A
Kilo Code	United States	Undisclosed	Acquired by Anaconda

Collaboration, enterprise knowledge and AI colleagues

These four products give AI systems access to context generated in daily work. Mio operates inside Slack, while Display.dev prepares documents for agents. HelloTwin.ai builds a semantic model of a business. Novum Studio turns team conversations into searchable knowledge through Ahoy. Their common challenge is access to context without forcing employees into another application.

Company	HQ Country	Value (US$)	Stage
Mio	France	$2.2m	Pre-seed
Display.dev	France	$0.51m	Pre-seed
HelloTwin.ai	Germany	Undisclosed	Investment
Novum Studio	United Kingdom	Undisclosed	Funding

Service providers

The service-provider deals focused on implementation capacity rather than new software products. VINCI Energies has signed a business combination agreement with All for One Group and announced a voluntary takeover offer. If completed, the deal would give its Axians business an established enterprise-applications and transformation platform in Europe.

Datapizza raised capital to expand its AI delivery work. SQUER’s reported majority investment backs a software-engineering consultancy. The activity suggests that delivery skills are becoming as scarce as some of the technology.

Company	HQ Country	Value (US$)	Stage
All for One Group	Germany	$370m	Takeover offer by VINCI Energies announced; completion pending
Datapizza	Italy	$11.2m	Series A
SQUER	Austria	Undisclosed	Private equity / reported majority investment

Beyond BARC’s core focus

The remaining 29 investments come from markets adjacent to BARC’s main coverage of data, analytics and AI management. They matter because compute capacity, model economics and vertical applications shape what enterprises can buy and operate, but they receive less analysis here.

Compute capacity and financing

Company	HQ Country	Value (US$)	Stage
CoreWeave	United States	$2.6bn	Debt financing
Verda	Finland	$25m	Debt financing
AI Infrastructure Capital	Switzerland	$18.6m	Launch round

Deployment and model delivery

Company	HQ Country	Value (US$)	Stage
SambaNova	United States	$1bn	Series F
Spectro Cloud	United States	$100m	Series D
Gradium	France	$30m	Seed extension
Fly.io	United States	$25m	Series D

Model development and inference economics

Company	HQ Country	Value (US$)	Stage
Fireworks AI	United States	$1.505bn	Series D
Together AI	United States	$800m	Series C
Multiverse Computing	Spain	$570m	Series C
Nous Research	United States	~$75m (reported)	Series B
Sherpa.ai	Spain	$18m	Growth round
kausable	Germany	$13.4m	Seed

Data access and storage

Company	HQ Country	Value (US$)	Stage
Oxylabs	Lithuania	$130m	Growth round
Tiger Technology	Bulgaria	$10m	Series A

Industrial AI and physical operations

Company	HQ Country	Value (US$)	Stage
InstaLILY	United States	$60m	Series B
Auger	United States	$50m	Series B
Arrakis	United Kingdom	$30m	Series A
Applied Computing	United Kingdom	$20m	Series A
Prodlane	Germany	$4.57m	Funding
Sophia	Germany	$1.1m	Pre-seed
SPREAD AI	Germany	Undisclosed	Strategic Series B extension

Customer service and commercial operations

Company	HQ Country	Value (US$)	Stage
Genius AI	United States	$44m	Series D
Cue	South Africa	$5m	Series A
Panora	France	$5m	Seed
Birdsview	Germany	$2.7m	Seed
Flyboard	Spain	$2.1m	Pre-seed

Domain-specific decision intelligence

Company	HQ Country	Value (US$)	Stage
Flow Labs	United States	$14m	Series A
Sensesbit	Spain	$1.1m	Seed, inferred

Who is doing the buying

Corporate investors usually look for more than a financial return. July produced three clear examples:

• Nvidia participated in Fireworks AI’s round. Greater use of the inference platform can increase demand for the compute on which it runs.
• Dell Technologies Capital and Intel Capital backed Fly.io, which provides infrastructure for agentic applications.
• Lockheed Martin Ventures invested in SPREAD AI, whose platform structures engineering data for industrial and defence projects.

Each investment links the parent company to a relevant workload or customer group. The startups gain capital and access to infrastructure, industry expertise or commercial channels.

The acquisitions in BARC’s main coverage follow five strategies:

• Progress’s agreed purchase of Domo’s assets would add an established BI and AI products business, along with its customers.
• NetApp’s purchase of DataPelago adds accelerated processing. Infoblox’s agreed acquisition of Kentik would extend network infrastructure into observability.
• Prefect’s acquisition of Dagster combines two workflow orchestrators under one owner.
• Vena’s agreed acquisition of Morpheo AI and Anaconda’s acquisition of Kilo Code add agentic capabilities to workflows where the buyers already have a strong position.
• Keyrus’s acquisition of inlumi and VINCI Energies’ takeover offer for All for One Group add implementation and transformation expertise rather than another software product.

Buyer and target geography reveal a clear contrast. The US deals concentrated on domestic product assets: Domo, DataPelago, Dagster, Kentik and Kilo Code strengthen platforms across analytics, processing, orchestration, observability and coding.

European activity showed a different form of regional consolidation: French buyers acquired or moved to acquire targets in the Netherlands and Germany. Keyrus acquired Dutch inlumi to expand its EPM and finance-transformation platform, while VINCI Energies launched a takeover offer for Germany’s All for One Group to combine SAP, cloud and AI expertise with Axians’ infrastructure, networking and cybersecurity capabilities.

The integration risk is clearest in the adjacent-layer acquisitions. Buyers still have to connect the products and teams into one offering rather than leave them as separate modules.

What we are watching next

• Enterprises will increasingly expect identity, policy, evidence and cost controls to work across models, applications and infrastructure rather than remain separate point products.
• Progress’s agreed Domo transaction will test whether an established BI platform can join a broader data and application portfolio without losing its identity or customer momentum.
• Credible Data and Morpheo AI point to demand for governed definitions and context shared by analytics, planning and AI agents.
• Further acquisitions around data processing, orchestration and observability look likely. Buyers will need to remove operational hand-offs, not merely expand their portfolios.
• Europe generated more primary events, but the United States captured more disclosed funding. The question is whether the UK’s larger governance and security rounds and Germany’s smaller finance and data deals can scale independently or become inputs to larger platforms and service groups.

Signals from related markets

• Capacity still attracts the biggest cheques, but cost pressure will favour providers that use it efficiently. Model compression, specialised inference and better workload placement all push in that direction.
• Vertical AI products will be judged on specific workflow results. The stronger offerings will use trusted data, act inside existing systems and produce results that people can verify.