Glasswing Ventures’ Enterprise AI Value Creation Framework
Not all AI-native products are created equal.
This is not a novel statement, but it is one that must be kept top-of-mind as one explores the ever-evolving Artificial Intelligence landscape. With the AI market expected to reach $2.7 trillion by 2032, the ability to discern which AI capabilities are most critical to various use cases will be necessary for adopting the most beneficial AI-native solutions for one’s business.
As AI-originalist investors, Glasswing Ventures has rigorously and diligently evaluated countless AI technologies for over a decade. We draw on our technical and operational backgrounds to analyze how AI models are structured, the foundations and quality of the data on which they are trained, and the resulting potential for meaningful business impact. We have crystallized our expertise and insights into a detailed framework we follow to help us identify future market shapers as they emerge.
Today, we are proud to unveil and open source Glasswing’s proprietary Enterprise AI Value Creation Framework. This framework is designed for AI pioneers and adopters — builders and buyers alike — to utilize in their evaluation, selection, deployment, and creation of AI products.
Depicted as a chart with an x- and y-axis, the Enterprise AI Value Creation Framework represents a comparative analytical model. The value creation potential of AI tools and platforms is positioned on the vertical axis (y-axis), while the criteria that determine the value are indicated on the x-axis.
As all AI applications should first be evaluated with their end-use in mind, we have designed the framework to represent a mathematical function encompassing aspects of utility theory, where the expected utility (or value) is maximized under given constraints and inputs. This is illustrated as the function f(x) of Data Quality and Propensity to Adopt, which integrates variables to predict the impact of AI solutions on the identified use cases.
To provide an example of how to read the Framework, we have included two use cases: AI in compound manufacturing processes (represented by light blue) and AI in content marketing (represented by dark blue). The area under the use case’s line indicates the potential impact the solution will have on an enterprise business. The varying heights of the line refer to the required access of the AI to different aspects of the business aspects as well as the associated risks of the technology.
Click the arrows to learn more about each criterion.
To say one has an AI company is a monolithic notion. Artificial Intelligence represents the what — an output of a much broader system of architectures, techniques, data sets, and training algorithms working together to produce the correct output for an appropriate use case. Glasswing’s Enterprise AI Adoption Framework deconstructs the three key criteria that determine the potential influence of an AI-native platform on an enterprise business: data, architecture, and impact.
Data: While data is the backbone of any successful AI model, simply “having data” will not guarantee an impactful model. Rather, the data must be clean, maintained, and relevant to the use case. Glasswing’s Framework outlines the elements one should look for in the data that an AI model is trained on to ensure the model leverages the right data in the right way for the appropriate use case.
Data Quality: At the most basic level, the higher the quality of data an AI model is trained on, the higher the quality of output the model will produce. While every AI model benefits from high-quality data, some use cases require higher-quality data than others. For example, AI models used in compound manufacturing must be trained on exceptionally high-quality data due to the severe consequences that result from an error. AI models used in content marketing, however, can be trained on lower-quality data because the repercussions of an error are less impactful to the overall business. In use cases involving lower-quality data, there must always be an option for a human to remain in the loop to review any output before it is published. This practice helps ensure accuracy and reliability, as humans can catch errors, biases, or inconsistencies that automated systems might overlook or misinterpret due to the poor quality of input data.
Data Availability: Data availability refers to the data to which an AI model has access. The specificity of the data directly impacts the specificity and security of the model. Scarce and private data refers to data that is unique to a business and difficult for the public to access, such as patient information. AI models used for compound manufacturing should be trained on scarce, private data to protect intellectual property and ensure compliance with the industry’s stringent regulatory standards, thereby maintaining high quality and minimizing risk.
Conversely, abundant and public data refers to data that is more widely available and is not necessarily specific to an individual business. AI-powered content marketing tools function well on abundant data due to the variety of answers one may look for when leveraging the product, as is the case with ChatGPT.
Fault Tolerance: Fault tolerance refers to a use case’s ability to handle or recover from an error made by the AI model. Content marketing has a higher fault tolerance than compound manufacturing due to the element of human-in-the-loop naturally found in its workflow as well as the lesser repercussions of errors.
Think of it this way: When one leverages a content marketing AI product to draft emails, they review (or should review) the email copy prior to hitting send to make sure the email is in their tone of voice and grammatically correct. If they don’t review the output, they risk sending an email with hallucinations, which will hurt their personal and professional brand reputation.
The risk of an AI model malfunctioning within the workflow of producing a pharmaceutical drug, on the other hand, could cost millions of dollars and threaten human lives. For this reason, there can be little to no margin of error. When exploring AI products for such important and sensitive use cases, one should look for models that incorporate features such as feedback loops, duplicated subsystems or components of systems, and that leverage Machine Learning for predictive maintenance to help close the margin of error.
Architecture: Where AI is the what, its architecture is the how. The architecture of an AI application refers to the compilation of models, guardrails, data, and systems that make it function. Understanding certain nuances of an AI model’s system, regulatory sensitivity, and model precision will help one make the right AI decision for their business.
System: AI systems can be broadly categorized as open or discrete, though in practice, they often exist along a spectrum rather than fitting neatly into one category. The choice between open and discrete systems depends on the specific requirements and risk tolerances of the use case.
Open-ended systems are characterized by the public accessibility of their components, such as data, code, and model structure. This transparency facilitates collaboration, allowing researchers, developers, and organizations to build upon each other’s work. However, this openness also introduces significant security and privacy risks, as sensitive information or vulnerabilities may be exposed to malicious actors. Open systems are particularly well-suited for use cases like content marketing, where models benefit from exposure to diverse data and require flexibility in handling a variety of prompts.
The models, data, and code used in discrete systems, on the other hand, are kept private. Discrete systems prioritize security and control, making them ideal for industries where confidentiality, safety, and precision are paramount, such as in compound manufacturing or aerospace.
Regulatory Sensitivity: As the phrase suggests, regulatory sensitivity refers to the extent to which an AI model complies with the existing regulations of its use case. While different regulations may apply to different use cases within one industry, such as the regulations of AI used on the manufacturing floor versus the AI used by the marketing team of a manufacturing company, all AI systems must operate within the legal frameworks of the industry.
For example, AI used to optimize production on a manufacturing floor must adhere to rigorous safety and operational standards, as any malfunction could directly impact product quality, worker safety, and compliance with industry certifications. Conversely, the AI used by the marketing team of that same manufacturing company operates under fewer constraints, as its outputs — such as campaign targeting or customer engagement — pose less risk to human safety or product integrity.
Despite these differences within use cases, all AI systems must navigate their industry’s overarching legal frameworks, which are often evolving to keep pace with technological innovation. In highly regulated environments, sensitivity to these frameworks is not just a compliance requirement; it’s a competitive advantage. Companies that design AI systems with regulatory foresight not only mitigate risks but also build trust and credibility in markets that are increasingly attuned to ethical and lawful AI deployment.
Model Precision: The precision of an AI model is calculated by the ratio of true positives (factually correct outputs) to the total number of positive predictions (outputs). This metric is a cornerstone for evaluating AI performance and should be tailored to the specific demands of each use case.
In high-stakes environments like compound manufacturing, precision isn't just a metric —it’s a mandate. AI systems predicting chemical formulations, monitoring quality, or ensuring consistency in production must adhere to extremely high precision benchmarks. A single false positive — such as approving a flawed product — can have cascading consequences, from operational inefficiencies to compromised safety. Here, precision is synonymous with trust, making near-perfect accuracy a non-negotiable.
Precision benchmarks in content marketing, on the other hand, are a bit more flexible. AI models generating ad copy, personalized recommendations, or campaign strategies don’t face life-or-death stakes. Rather, the focus of these models is on creativity, engagement, and relevance. A slightly off-target suggestion or a less-than-perfect phrase might not be the ideal output, but it’s unlikely to cause significant harm. Precision in marketing supports optimization rather than absolutes.
Impact: While understanding the data and architecture of AI models will provide one with the technological know-how required to choose the right AI solution for their business, it is equally important for one to understand the strategic business impact of the AI product.
As adopting and integrating AI is a costly and resource-intensive project, understanding the anticipated impact of AI adoption upfront will be essential to ensuring one allocates resources to the project that brings the highest return on investment. While impact does relate to revenue, it also includes the ease of workflow integration and the propensity for teams to adopt the technology.
Workflow Integration: Adopting any AI technology disrupts the established workflow of a business function. For instance, if the purpose of adopting AI is to replace an existing tool in one’s tech stack, then careful planning is required to not only decommission the old system but also to ensure that the new AI solution seamlessly communicates with the technologies already in place. Doing so could involve significant backend integration efforts as well as user training to close any knowledge or functionality gaps.
Alternatively, if AI supplants manual tasks previously handled by employees, a strategic realignment of roles is required. Decision-makers must redefine employees’ roles to leverage their human skills in areas where AI cannot perform efficiently.
The ease of AI integration varies significantly by use case. AI solutions in content marketing have gained rapid traction due to their relatively straightforward integration into existing workflows. These systems are often user-friendly, require minimal adjustments to current processes, and can dramatically enhance efficiency and creativity in content generation. AI implementations in industries like compound manufacturing, on the other hand, demand extensive planning, a deep understanding of existing processes, and significant resources to be integrated effectively.
Revenue Impact: The impact an AI platform has on a business’s revenue can be either direct or indirect. When an AI model has direct revenue impact, the impact is concrete and measurable, often reflected in metrics such as increased customer acquisition, heightened product sales, reduced customer churn, or significant cost reductions. AI models providing indirect revenue impact might not have immediate measurable results in financial statements but are crucial for long-term growth.
In specialized fields like compound manufacturing, AI-native solutions often yield a direct impact on revenue by optimizing production processes and reducing the frequency of costly errors. Glasswing’s portfolio company, BaseTwo, for example, has built a platform that helps its customers see a 20% reduction in process deviations.
Conversely, AI applications in content marketing typically deliver an indirect impact on revenue. These tools refine content strategies and enhance user engagement, gradually building a stronger brand presence for companies. Although these benefits might not immediately and directly impact revenue figures, they lay the groundwork for sustained growth by fostering a loyal customer base and improving one’s market position.
Propensity to Adopt: The likelihood for teams to adopt an AI platform directly ties to the two previously described impact criteria. If employees are able to clearly understand how an AI solution will improve their workflows and the overall business, they will be more willing to accept change.
Despite the significant effort required to integrate AI systems into compound manufacturing, the potential for the solutions to have a transformative business impact makes the initial challenges worthwhile. Because employees in this sector can often see the direct correlation between AI implementation and improved operational outcomes, they are more willing to engage with new technologies.
While AI solutions for content marketing are simpler to deploy and mesh into existing workflows, the weaker ties to revenue lead to lower motivation among the broader enterprise to fully and completely adopt these products. With such a use case, it is evident that the ease of integration alone does not guarantee widespread acceptance of AI technology unless the impact on key performance indicators is evident and aligned with strategic business goals.
The Glasswing Perspective
The current AI landscape is vast, and it will only continue to expand in the coming years. By taking the time now to understand the essential criteria for AI-native platforms built for the enterprise, one will position themselves to adopt the right AI solution for their business and use case.
By following Glasswing’s Enterprise AI Adoption Framework and understanding the data, architecture, and impact of an AI model, AI pioneers will feel empowered to make AI decisions that lead to positive business outcomes. If you have any questions or comments about the Framework, please do not hesitate to reach out to info@glasswing.vc.
