CB Insights releases its annual selection of the world’s most promising artificial intelligence (AI) startups, highlighting the top emerging private AI companies based on market traction, quality of investors, and talent.
This year’s 100 Ventures highlights the AI industry’s focus on early growth and market validation. Analyzing the situation, we see that Series A venture companies are the most prominent group on the list, containing a total of 60 companies.
These companies have gone beyond the prototype stage and proven that their products solve real problems. They are now focused on scaling customer acquisition, turning early traction into a repeatable business model, and building their sales and marketing team.
These companies include 7AI, DeepJudge, and StackOne. 7AI is a US-based company that deploys fleets of autonomous AI agents to automate security alert processing, incident management, and threat hunting.
Founded by former Cybereason CEO Lior Div and backed by Greylock Partners and Index Ventures, 7AI secured a US$130 million Series A in December 2025, just 10 months after starting from stealth. The company plans to use the proceeds to expand its AI security engineering and go-to-market teams. In 2025, its platform reportedly processed over 2.5 million alerts and 650,000 investigations, reduced investigation time from hours to minutes, and eliminated up to 95-99% of false positives in production.
Switzerland’s DeepJudge provides an AI-powered knowledge search and document retrieval platform that connects law firms’ document management systems to search-enhanced AI agents. Founded by a former Google researcher with a PhD in AI from ETH Zurich, the company serves elite companies such as Freshfields and Holland & Knight, and has achieved active usage rates of 80-90%.
DeepJudge secured US$41.2 million Series A in November 2025 to deepen product innovation, grow its partner ecosystem, and expand internationally.
UK-based StackOne builds unified APIs and tool invocation interfaces that connect AI agents and Software-as-a-Service (SaaS) applications to more than 200 enterprise systems across HR, CRM, and collaboration tools.
Founded by former Google and Oracle executives, StackOne raised $20 million in Series A in May 2025 to continue building the most advanced tool-invoking large-scale language models (LLMs), invest in research and development (R&A), and further expand the number of integrations and depth of actions available on the StackOne platform.
Early stage innovators and scaling leaders
Following the Series A stage leaders, 21 seed stage ventures will be featured. These young companies are still refining their early products and finding products that fit the market.
These include Voltz, a US startup that develops AI models for the design and optimization of molecular structures, and Paid, a UK-based billing infrastructure provider for AI agents.
This year, Volz secured a multi-year collaboration with Pfizer and was backed by Andreessen Horowitz alongside Zetta Venture Partners and Amplify Partners. Backed by Lightspeed Venture Partners and Sequoia Capital, Paid has secured high-profile enterprise customers such as IFS and Artisan, and has grown its headcount by 140% year-over-year (YoY), according to CB Insights.
Meanwhile, 16 companies are at Series B or higher stage. These organizations have established business models and are experiencing significant revenue growth. They are currently expanding internationally, expanding their team, and actively acquiring customers.
These include Serval, which provides an AI-powered IT service management platform that automates help desk requests; Actively AI deploys resident agents to automate sales accounts. Positron develops purpose-built AI inference hardware optimized for transformer models.
Serval serves clients like Perplexity and enables over 50% ticket automation. According to CB Insights, the company grew its workforce by 347% year-over-year with support from Sequoia and Redpoint Ventures.
Founded by Stanford AI researchers, Actively AI drives measurable ROI, enabling clients like Samsara to double their conversion rates. Actively AI grew its workforce by 107% year-over-year and achieved top 4% hiring momentum.
Finally, Positron, which makes chips in Arizona, claims 3.5 times better performance per dollar than Nvidia’s H100, with memory bandwidth utilization of 93% versus 10-30% for GPU systems. The company grew its workforce by 120% year over year.
Nine startups specializing in financial services made the list
This year’s CB Insights AI 100 list spans three broad categories:
- Enterprise application provider. We offer AI products that deploy agents and workflows to automate or enhance business functions across your organization.
- Industry application provider. We offer verticalized AI products built for your industry’s specific data, compliance, requirements, and workflows. and
- Infrastructure and compute provides a foundational layer of models, tools, hardware, and observability on which to build AI applications.
Within the industrial applications category, financial services emerged as the largest subsector, with nine companies participating. This number is consistent with medical care.
The nine AI companies focused on financial services are:
- Agent Smyth is an autonomous agent platform for trading and investing based in New York.
- Avantos is an AI-native operating system for customer management that enables financial institutions to onboard, service, and grow customer relationships.
- Bretton AI, an AI agent platform for financial crime compliance.
- Casap is an award-winning dispute automation platform that streamlines compliance, reduces operational costs, and increases consumer satisfaction for banks, credit unions, and fintech startups.
- FutureAI is domain-specific AI for the insurance industry that automates workflows such as application intake, policy comparison, and underwriting processes.
- Light is an AI-native finance platform that automates payments, expense management, bookkeeping, and financial reporting across jurisdictions.
- Questflow is a collaborative AI automation tool for teams to create, orchestrate, and automate tasks across platforms.
- Salient is an AI workflow automation tool for lenders that uses compliance-first AI to automate collections, customer service, disputes, chargebacks, and total loss mitigation. and
- Tidalwave is an agent AI company that aims to eliminate bottlenecks in the mortgage process.
Physical AI becomes a separate category
A defining trend of this year’s AI 100 list is the debut of physical AI as a standalone category. Physical AI refers to AI systems that perceive, make decisions, and act on the physical world through autonomous machines and robotics. This sector has seen significant momentum over the past few years, with companies in this sector raising a record USD 78 billion in 2025.
Eleven companies make up this new category, spanning robotics software, autonomous hardware, and enabling chips. The participation of these companies reflects a significant inflection point in physical AI, as the entire stack required to deploy autonomous systems in the real world is maturing at the same time.
Beyond these capabilities, the field is also seeing a proliferation of solutions that coordinate multiple autonomous units toward a common goal. The key players illustrating this change are:
- InOrbit builds AI-powered software to monitor, coordinate, and optimize fleets of autonomous robots in warehouses, factories, and other industrial environments.
- FieldAI develops AI systems that enable various types of robots to operate autonomously in complex real-world environments such as construction, energy, and logistics sites. and
- Gravis Robotics brings AI and autonomous technology to excavators and other heavy construction equipment to improve productivity, safety, and accuracy on job sites.
According to CB Insights, InOrbit has grown its customer base by 200% in the past year. FieldAI raised $314 million in Series A in August 2025 at a valuation of $2 billion. Founded in 2022, Gravis Robotics has already expanded to seven countries, showing early signs of a rapidly forming market.
The rise of agent AI brings observability and evaluation tools
Another distinctive trend this year is the evolution of AI agents into a separate actor class. As these agents move from experimental prototypes to active participants in enterprise workflows, they require their own layers of identity, authentication, and accountability.
The observability and evaluation category of this year’s AI 100 list addresses this. Nine companies in this category have built the necessary infrastructure to manage these non-human actors and established operational rulebooks for bringing autonomous systems into enterprise workflows. These companies include:
- Keycard builds identity and access management infrastructure that enables enterprises to securely authenticate, authorize, and manage AI agents.
- Geordie AI provides governance and security oversight for enterprise AI agents, giving teams visibility into how agents are performing and what risks they pose.
- Virtue AI develops AI security software that red-teams, monitors, and enforces guardrails on AI models and autonomous agents. and
- Striker protects AI agents at runtime by detecting vulnerabilities, testing agents against attacks, and preventing malicious AI behavior on enterprise systems.
Nine companies in the observability and valuation category have raised a total of US$278 million over the past three years. This influx of capital highlights the nascent stage of this category and the growing recognition that robust governance and oversight are critical to the secure integration of autonomous systems into enterprise workflows.

Featured Image: Compiled by Fintech News Switzerland Based on image by arslantanoli via Magnific
