All businesses confront an escalating challenge. This challenge involves the ability to leverage data science to create organizations that are directed by more reliable decision-making.
As newly generated enterprise data increases exponentially, the availability of data science skills will only increase linearly, at best. This mismatch will further hinder an organization’s ability to derive value from its data.
Furthermore, our research suggests that 70% of corporate data remains inaccessible, unanalyzed, or dormant, indicating that substantial value goes unrealized. Consequently, only 4 in 10 business leaders trust their data-driven decisions. It is only going to get worse.
This research note examines how AI data science agents that work alongside human co-workers represent the only feasible path forward.
We will also preview the upcoming AI for Data Scientists Conference (cAI25), hosted by causaLens. This “must-attend” event aims to assist enterprises in addressing their data science challenges with innovative agentic AI systems technologies and best practices learned from industry pioneers.
Finally, we will explore the key factors that will make employing AI data science agents inevitable for most enterprises as well as how causaLens is already assisting enterprises to revolutionize their data science capabilities.
Why Attend cAI25
The AI for Data Science Conference (cAI25) will take place in San Francisco on April 1st, 2025, with a global livestream audience. This conference offers a unique opportunity to explore how cutting-edge AI and enterprise data science innovations will shape the future of intelligent AI agents and agentic workflows.
At the conference, attendees will learn agents can transform their enterprise data science teams and empower an extended community of “citizen data scientists” across their organization.
Through exclusive interviews and in-depth discussions, participants will learn how leading organizations integrate AI agents into their workflows to enhance productivity and improve decision-making, even amid the growing data science and AI talent gap. Industry leaders will also share strategies for automating complex analytics workflows with autonomous systems, allowing teams to scale operations efficiently and swiftly adapt to new opportunities.
Another key theme at cAI25 will be how organizations can put their decision-making on autopilot using AI data science agents powered by technological advancements in generative, semantic, and causal reasoning. This includes best practices for delivering transparent and explainable outcomes, key to creating trust across an organization and ensuring regulatory and policy compliance.
Overall, this conference presents a unique opportunity to explore the future of enterprise decision-making and examine how visionaries are already delivering next-generation decision-making platforms, enhanced by advanced AI agents, to facilitate more informed and strategic decisions.
The cAI25 Experience
From theCUBE Research’s participation in previous cAI conferences, we regard the networking opportunities as one of the most compelling reasons to attend in person rather than rely on the live stream broadcast. These conferences are strategically designed to foster meaningful connections among industry leaders, innovative companies, and AI experts.
The conference provides a variety of interactive discussions, collaborative sessions, and informal gatherings that promote knowledge exchange, relationship-building, and idea-sharing. The conference atmosphere fosters dialogue, enabling attendees to form valuable connections, gain insights into emerging industry trends, and explore practical collaborations.
You will have a unique opportunity to engage directly with professionals at the forefront of AI-driven transformation across various industries, especially those well ahead of the agentic AI adoption curve. The primary goal of the conference is for attendees to leave with practical knowledge, strategies, and connections that can significantly enhance their organizations’ AI and data science capabilities.
For instance, you will gain unique insights from Guido Imbens, a Nobel Prize-winning economist at Stanford University renowned for his contributions to causal inference. His groundbreaking methods for analyzing cause-and-effect relationships, such as natural experiments, have significantly influenced econometrics, statistics, and advancements in AI.
You will also gain a unique opportunity to engage directly with those implementing real-world AI decision intelligence solutions from leading brands such as Meta, GE Health, Bosch, and Johnson & Johnson.
By attending this event, you’ll do more than expand your network — you’ll become part of a vibrant community committed to shaping the future of AI decision-making. We at theCUBE Research believe this represents the next frontier in AI.
The agenda will concentrate on delivering real-world insights regarding the progressive integration of AI agents into enterprise data science workflows, which includes:
- Keynotes from industry leaders
- Panel discussions and fireside Chats with experts
- Hands-on workshops to build knowledge and skills
- Real-world case studies and demonstrations
You will first learn how AI agents can scale your enterprise data science capabilities and automate complex analytics workflows to uncover previously hidden insights, ultimately improving data-driven outcomes. This approach allows you to build significantly more productive and resilient data science teams without substantially expanding your human resources.
Secondly, you will discover the strategies that pioneering companies utilize to enhance AI trust, transparency, and policy compliance while balancing innovation at scale and with speed.
Finally, you will experience the innovation of next-generation AI enterprise decision-making platforms powered by advanced AI agents, agentic workflows, and groundbreaking technologies like causal discovery and inferencing.
Whether you aim to scale your analytics across the enterprise, automate tedious data processes (such as data cleaning, text-to-SQL, and model building), implement advanced AI decision-making systems, or enhance the value of your data science teams, this conference is ideal.
You can register here: cai-conference@causalens.com
Why AI Data Science Agents
What if everyone, from marketing, operations, and sales to R&D, finance, and legal, could self-apply advanced data science and AI techniques to leverage one of their most valuable resources — their data?
And be empowered to explore countless possible futures by evaluating the consequences of potential decisions and understanding the specific chain of events that lead to a desired outcome, with trust and confidence. In other words, understand WHAT to do, HOW to do it, and WHY certain actions are better than others.
Almost every business would say, “Yes, please, now!”
However, to do so, enterprises must first overcome three key realities:
- The inability to hire or afford enough data scientists.
- Data is growing faster than businesses can keep up.
- A growing lack of trust in data & AI outcomes.
Let’s explore each of these realities in more detail, which we believe will make the case for pursuing a totally new approach to enterprise data science. An approach anchored in the power of augmenting human talent with agentic AI systems to simplify and automate data science for everyone across an enterprise.
The Data Science Talent Crisis
As data science and AI adoption accelerate, enterprises are facing an unprecedented AI talent crisis — one that is expected to intensify in the years ahead. Consequently, companies must fundamentally rethink how they develop, acquire, and maintain AI expertise to stay competitive.
The statistics should concern all business leaders with AI aspirations. The demand for data science and AI professionals continues to outpace supply — by nearly 3x. Meanwhile, the number of new AI-skilled entrants into the workforce lags 10x behind new AI job openings. According to a recent Boston Consulting Group study, enterprises are already feeling the strain: 7 in 10 enterprises are struggling to find data science and AI talent. This is further supported by a March 2025 study by McKinsey & Company illustrating that data science skills are one of the most difficult to secure among all data and AI-related professions.
Another concern is the cost of acquiring and retaining these skills. Multiple surveys indicate that almost half of those employed are contemplating leaving their current jobs for better-paying opportunities. This expanding skills gap is driving an increase in AI salaries, with some companies paying 47% more to secure top AI talent. As reported by Menlo Partners, some pay 3x salary premiums and offer massive equity packages to ensure the best of the best.
Furthermore, it is estimated that over 200,000 data science and AI job openings will be created in the U.S. alone by 2025, making full-time hiring an impractical strategy for many enterprises.
As enterprises face these challenges, they increasingly recognize that sourcing enough human talent is an unrealistic strategy.
This is why theCUBE Research believes that AI talent augmentation will no longer be optional — it will be essential. Companies that combine on-demand AI human talent with AI-powered agentic systems will more effectively navigate the AI skills shortage and empower a wide range of their employees with data science skills. Organizations that embrace this shift will benefit from the momentum, while those that hesitate will face greater challenges in securing and retaining AI talent.
The Data Growth Dilemma
While we are confident that most businesses will conclude that they cannot hire or afford enough data scientists to support their needs, there is more to this case for employing AI data science agents and agentic workflows.
Data is growing faster than businesses can keep up. Currently, 70% of enterprise data is either inaccessible or unanalyzed, and this problem will worsen exponentially over time.
Over the next five years, enterprise data is expected to grow exponentially, driven by the proliferation of digital interactions, connected devices, AI systems, and real-time business processes. By 2030, the volume of newly created enterprise data could reach 400 zettabytes — a nearly fivefold increase from the projected 2025 growth in new data.
Unfortunately, the vast majority of this new data will remain unanalyzed without a fundamental change in approach. The challenge isn’t merely about collecting data — it’s about transforming that data into insights. That’s where the bottleneck is increasingly out of reach for most enterprises.
Many business leaders will intuitively look to scale their armies of data scientists as the answer. However, the harsh reality is that enterprises cannot scale enough of this talent to keep up with the exponential growth of newly generated data.
For example, even if an enterprise grows data scientists linearly— say from the equivalent of 30 to 60 “zettabytes of analytical power” over the same period — the gap between the data being analyzed and the data being generated will widen significantly.
This equates to a rapid decline in the percentage of newly created data being analyzed, from an estimated 36% today to a mere 16% by 2030.
To address this imbalance, enterprises must rethink their data strategy. This includes investing in automation, AI agents, and next-generation data science platforms — tools that can amplify the productivity of existing talent and bring more “citizen data scientists” into the fold to handle this unprecedented growth in data.
The goal should not be to endlessly hire more data scientists— rather, it should be to make data science accessible to everyone across a business while using AI data science platforms to automate what humans are incapable of doing.
The Trust Factor
If the shortage of affordable skills and the diminishing ability to analyze newly generated data is not challenging enough, let’s take into account the trust factor.
A significant consequence of the growing gap in unanalyzed data will be a steady decline in business leaders’ trust in data and AI-driven recommendations. Currently, only about 38% of executives report trusting their data and AI analytics when making decisions — and that percentage is projected to drop by 20 points to just 18% by 2030 if current trends continue.
This erosion of trust stems from a mounting frustration with incomplete, outdated, or unanalyzed data that undermines decision-making. As enterprise data grows exponentially while analytical capacity lags behind, leaders increasingly confront noisy signals, inconsistent insights, and dashboards that fail to reflect business reality.
In addition, as enterprises pursue AI-driven automation and data science, a significant roadblock is emerging — the trust factor.
As the McKinsey & Company 2024 State of AI study reported, over 60% see inaccuracies, bias, and the concealment of influential factors in AI outcomes compounded by the lack of explainability as a key inhibitor to using AI in decision-making use cases. Bain & Company has also echoed these concerns, reporting that inaccuracies, regulatory concerns, and data privacy were impeding trust in AI decision-making.
Going forward, trust will become the currency of innovation. The more AI is trusted, the more businesses will be willing to “buy into” the innovation promised by Agentic AI.
The capability of a new generation of data science platforms built as agentic systems promises to generate answers and to explain how and why they arrived at those answers, which will become the new standard for enterprise-grade data science. Companies prioritizing AI transparency will gain a competitive advantage, while those that do not will face increasing resistance from regulators, customers, and internal stakeholders.
Revolutionizing Data Science
causaLens, a London-based startup, is on a transformative mission to revolutionize enterprise decision-making by introducing AI Data Scientists — advanced AI agents that work seamlessly together and alongside human teammates. Their solutions empower organizations to fully harness their data by automating complex data, AI and analytical tasks traditionally performed by human data scientists.
This novel approach enables enterprises to overcome the limitations of traditional data science workflows and their reliance on human talent. It significantly enhances enterprise data science processes, achieving greater scalability, efficiency, speed, trust, and analytical depth with specialized AI agents that can independently manage data cleaning, exploratory analysis, predictive modeling, and even advanced causal inference.
As Darko Matovski, the CEO and co-founder of causaLens, stated:
“We’re moving beyond dashboards and co-pilots to where AI agents are becoming real teammates to create auto-scalable, intelligent decision-making engines that will revolutionize enterprise data science.”
Real-world deployments by Fortune 100 companies demonstrate the platform’s transformative potential. The platform achieves results at scale in weeks, which would have previously taken human teams months or longer. For example, a global enterprise successfully deployed 50,000 custom models within one week — a task that traditionally takes months.
This vision reflects that of industry luminaries such as OpenAI CEO Sam Altman, who predicts that AI agents will “materially change the output of companies” by joining the workforce this year. Meanwhile, NVIDIA CEO Jensen Huang envisions a world where IT departments evolve into “HR departments of AI agents,” managing and nurturing digital workers alongside human employees.
The broader implications of causaLens’ success are profound. It potentially democratizes AI and data science capabilities once limited to resource-rich enterprises. By providing unlimited AI data science resources for organizations of all sizes, causaLens empowers human data scientists with new “superpowers” and enables non-technical workers to engage in self-service data science activities.
With a leadership team that possesses extensive experience in applying AI to critical, high-stakes decision-making environments, causaLens is uniquely positioned to guide enterprises through this pivotal transition. The company’s vision is not just about improving productivity — it aims to fundamentally reshape the data science labor market and usher in an era of intelligent, real-time decision-making on an unprecedented scale.
The causaLens Agent Platform
causaLens’ AI Agent platform includes a range of specialized AI agents that collaborate within an agentic workflow to autonomously perform data science tasks, from data cleaning to advanced analytics and predictive modeling. Leveraging machine learning and advanced AI techniques, including causal AI, these agents can analyze data, identify patterns, and uncover drivers of outcomes to produce actionable insights without ongoing human oversight. Their main goal is to intelligently optimize the data science workflow, enabling organizations to fully leverage their human and data resources.
The platform facilitates seamless interaction between users and AI agents to ensure that appropriate agents address specific tasks within a managed workflow. Users maintain control over automation and can pause, edit, or rerun workflows as needed, fostering a collaborative environment that enhances trust, transparency, and explainability. It also enables seamless connections to leading enterprise data, AI, and cloud systems while enforcing security and compliance standards. This makes the platform adaptable to various deployment needs, designed to fit seamlessly into existing enterprise environments, whether on-premises or in the cloud.
We are also encouraged by their new collaboration with Google to integrate their AI data science agents and causal reasoning technologies with Google Cloud’s powerful AI infrastructure, including its Gemini models. This collaboration aims to enable large language models (LLMs) to analyze and reason about complex quantitative data, delivering a clear differentiator in enterprise-grade decision-making and problem-solving. These advancements suggest that the future of AI resides in multi-agent systems capable of processing and analyzing vast volumes of data and understanding and acting on underlying causes.
At the heart of causaLens’ approach is a profound commitment to trust, underscored by the belief that “automation without trust is irrelevant.” To ensure reliability, causaLens integrates causal reasoning and business-specific grounding into its AI agents, complemented by reinforcement learning and human oversight. Additionally, the agents adapt dynamically, retaining knowledge of organizational practices, compliance standards, and domain-specific methods, thereby continuously enhancing their relevance and effectiveness.
Within the platform, causaLens provides an array of out-of-the-box agents equipped with valuable skills, including:
- Planning Agent: Decomposes complex questions into data science workflows, selecting the optimal agent for each stage.
- Data Cleaning Agent: Automates multi-step data cleaning processes, ensuring high-quality datasets.
- Exploratory Data Analysis (EDA) Agent: Assists in data exploration and visualization to uncover underlyin, hidden patterns.
- Causal AI Agent (CAIA): Specializes in causal reasoning, automatically selecting appropriate models to determine causal effects within data.
- SQL Agent: Writes SQL queries to extract data from various sources, streamlining data retrieval processes.
- App Building Agent: Constructs interactive application components facilitating the sharing of insights with stakeholders.
causaLens has also indicated that they plan to deliver additional pre-built AI agents to enhance the platform’s capabilities.
In addition to these pre-built agents, the platform allows organizations to create custom agents tailored to specific business domains, knowledge sets, compliance policies, and organizational needs. Building your own custom data science agent can significantly enhance your organization’s analytical capabilities, enabling highly domain-specific decision-making. The causaLens platform offers a streamlined process to develop these agents, within a step-by-step agentic workflow:
1. Define Your Agent’s Objective — Define the goal of your agent, whether it’s to discover causal business drivers or to forecast future sales.
2. Create Your Agent’s Persona — Establish your agent’s focus, tone, and language model to align the agent to your unique environment.
3. Equip with Tools and Knowledge — Enable the agent to reason and problem-solve within its domain (RAG, Knowledge Graphs, Causal AI, etc.).
4. Train and Evaluate — Develop thorough evaluation and QM pipelines to ensure consistent performance across successive versions of the agent.
5. Publish to Users — Deployment to teams across your enterprise to engage with the agent, empowering them with data science “superpowers”.
This agentic workflow, along with human expertise, can radically speed the delivery of highly trusted data science agents tailored to any organization’s needs, enhancing their efficiency and decision-making capabilities.
The causaLens AI Agent platform represents a paradigm shift that will significantly reshape the role of human data scientists, enabling them to now concentrate on more strategic, complex problem-solving initiatives that leverage human creativity and domain expertise. Moreover, by democratizing advanced AI capabilities with an intuitive, conversational-like experience, everyone across an organization — from data analysts to strategic business leaders — can be empowered with new data science superpowers to make better, more informed decisions.
Real-world Impact
causaLens has demonstrated significant success across various industries by deploying AI Data Scientists, leading to substantial operational improvements and financial gains for their clients.
For instance, a leading oil services company realized an annual value of $30 million by reducing downtime through the implementation of a digital twin for its drilling systems. This solution allowed the company to comprehend the trade-offs between drilling speed and equipment failure rates, thereby facilitating more informed operational decisions.
A $1 billion industrial machinery manufacturer experienced a 10% reduction in order delays in the manufacturing sector. By integrating causaLens’s causal AI agent, the company identified root causes of delays, forecasted the likelihood of future delays, and received optimal recommendations for order allocation across manufacturing plants.
The financial services industry has also benefited from causaLens’s solutions. A large European bank with over €400 billion in aggregate deposits projected a return of €12 million by utilizing causaLens solutions to understand better and manage deposit risk. The platform gave the bank enhanced insights into deposit behaviors under changing macroeconomic conditions, improving risk management strategies.
Furthermore, a leading global insurer realized multi-million dollar reductions in auto claim costs by utilizing causal AI to analyze and optimize their claims processes. The insurer successfully quantified the impact of repairs performed within their network, resulting in more effective cost-reduction policies and enhanced decision-making concerning garage partnerships and customer incentives.
These examples illustrate the versatility and effectiveness of causaLens’s AI-driven solutions in addressing complex challenges across diverse industries. For more case studies, visit the cauaLens resources page.
What To DO and When
Enterprises face an intensifying challenge as the exponential growth of data increasingly outpaces the availability of skilled data science talent, it’s our view that organizations must urgently adopt innovative solutions to close this widening gap.
causaLens is addressing this critical challenge through its revolutionary AI Data Science Agent platform, allowing enterprises to scale their AI and analytics capabilities without proportionately expanding their data science teams.
The upcoming cAI25 AI for Data Science Conference, provides a unique forum for exploring these advanced agentic systems and to gain insights into real-world strategies for integrating AI agents into enterprise workflows.
We’d recommend the following actions to learn more:
- Attend the cAI25 conference
- Book a 15-minute demonstration
- Evaluate potential new use cases
- Explore the causaLens website
As always, let us know if we can help you!
We also recommend subscribing to the Next Frontiers of AI Podcast on the SiliconANGLE Media YouTube Channel or on Spotify to hear from industry experts and pioneers on what is shaping the future of AI in business.
Thanks for reading. Feedback is always appreciated.
