In the world of AI development, staying ahead of the curve is a constant challenge.

In early 2024, we started to build the AI Strategist, a tool designed to help businesses develop their own AI strategies without the need for deep technical expertise. As more companies became aware of the potential of AI but lacked the knowledge to implement it effectively, we saw an opportunity to productize our consulting process, offering AI-driven insights to help businesses shape their AI roadmaps.

What started as an innovative response to a clear market need ended with an unexpected pivot. What have we learned and how this experience changed our approach? Let’s delve into this.

Origins and Vision: AI as a Strategic Tool

The concept for AI Strategist emerged from our work with business founders and executives who sought guidance on how to integrate AI into their organizations. As artificial intelligence gained traction, we noticed a growing demand for strategic advice on how to apply AI effectively. Many business leaders were aware that AI could transform their industries but didn’t have a clear pathway to get started.

At Fractal Labs, we had been offering ad-hoc advice to individuals and companies on how they could leverage AI in their workflows and daily operations. However, it became clear that this consulting model wasn’t scalable. Clients wanted faster, more accessible ways to create AI strategies without the need for long workshops or extensive one-on-one sessions.

This led to the idea of AI Strategist, a productized version of the strategic consulting process. Our vision was to create a tool that allowed users to interact with an AI-driven system that could simulate the kind of strategic discussions we had been facilitating in person. By inputting their business challenges and goals, users could receive tailored insights, helping them to design actionable AI strategies. The platform would be especially useful for companies without internal AI expertise, enabling them to take the first steps toward AI adoption on their own.

From Concept to MVP: Building the AI Strategist

With a clear vision, we moved swiftly from concept to development. Our goal was to develop a Minimum Viable Product (MVP) that would validate the core idea: that an AI-driven system could guide users through the strategic decision-making process and help them build their own AI roadmaps.

In just a few weeks, we had a working MVP. The tool featured a conversational chat interface where users could interact with AI by answering a series of prompts designed to extract key business information. This conversation wasn’t just about inputting data—it was designed to mirror the kind of consultative questioning we used in our workshops. The AI would use the information to generate strategic recommendations, offering users insights into how AI could be applied to their specific business challenges.

Key Features of AI Strategist

• Conversational Interface: Users could engage with the system through a chat-based interaction. The AI would prompt users with questions, gather insights, and provide strategic guidance in real time. This conversational model was designed to simplify complex AI strategies, making them accessible to business leaders without technical backgrounds.
• Document Generation: We wanted our users to be able to create useful structured documents directly from their conversations with AI Strategist. As users interacted with AI Strategist, the system would compile their input and the AI’s recommendations into a formal strategy document. Users could adjust these documents as necessary, but would end up with a well-informed, ready-to-use AI strategy in hand.
• Predefined Templates and Prompts: The system was built with customizable templates that adapted to different industries and business types. These prompts helped guide users through the strategic process, ensuring that the final output was relevant to their specific needs.

The MVP demonstrated how AI could simulate the strategic thinking process, offering users a tool to explore AI applications within their business. It was built to be scalable, with the ability to handle multiple use cases and adapt to the unique challenges of various industries.

The Pivot: Facing Competition and Realigning Strategy

As we prepared to refine AI Strategist and bring it to market, we faced an unexpected shift in the landscape. Anthropic’s Claude AI, a major player in the AI field, launched a feature called Claude Artifacts, which closely mirrored what we were building with AI Strategist. It offered a polished interface and broader functionality, allowing businesses to interact with AI to generate strategic documents and store them seamlessly within the system.

It became apparent that Claude Artifacts had already delivered on many of the features we were working on, backed by the resources of a large AI company. After testing Claude Artifacts, we recognized that it was a highly competitive product that accomplished what we aimed to build—only with greater resources and a more advanced user experience.

At this point, we had a choice: continue investing in AI Strategist and attempt to compete with a well-funded player or pivot and focus our resources on new opportunities. After internal discussions, we made the strategic decision to put AI Strategist on hold. The timing was critical—we had built a solid proof of concept but had not yet heavily invested in scaling or marketing the product. This allowed us to pivot without sinking unnecessary resources into a project that would face stiff competition.

Lessons Learned: Adapting to the Market

The development and subsequent pivot of AI Strategist provided several valuable lessons for us at Fractal Labs:

1. Agility in Decision-Making: One of the key lessons was the importance of staying agile in a fast-moving landscape. The AI industry is constantly evolving, and being able to pivot quickly when new technologies emerge is essential. By choosing to refocus early, we avoided overcommitting resources to a project that no longer aligned with market realities.
2. Vertical vs. Horizontal AI Solutions: The experience reinforced the importance of focusing on niche, vertical AI solutions rather than broad, horizontal applications. Large players like Anthropic are well-positioned to dominate general AI tools, but smaller companies like ours can thrive by targeting specific industries or use cases. AI Strategist was designed to be versatile, but we realized that specialization might offer a more defensible position in the AI market.
3. Embracing Non-Attachment: Building successful products requires honesty with ourselves and a level of objectivity.We were excited about AI Strategist and saw its potential, but we understood that sticking with a project just because we had invested time and effort wasn’t the right approach. The ability to step back, assess the market, and pivot without emotional attachment was crucial to making the right decision.

Technical Gains: Mastering RAG and Embeddings

Despite shelving the AI Strategist project, the development process yielded significant technical advancements, particularly in retrieval-augmented generation (RAG) and embeddings. These technologies have since become part of our core capabilities, allowing us to apply them to future projects.

• Retrieval-Augmented Generation (RAG): This technique allowed us to retrieve relevant information from large datasets during AI interactions, ensuring that the AI responses were highly informed. Rather than relying on a limited context window, RAG enabled the AI to pull from a much larger knowledge base, augmenting the responses with specific, relevant data points.
• Embeddings: Our work with embeddings enabled us to store large volumes of data in vector databases, which the AI could then query for relevant information. This allowed us to handle vast datasets, making the AI system more intelligent and responsive to nuanced user queries.

These technical skills have already been incorporated into other projects, particularly those requiring complex data retrieval and analysis. The expertise we gained in RAG and embeddings will be invaluable as we continue to build AI systems that offer more specialized, data-driven insights.

The Future: Reapplying AI Strategist Technology to New Projects

The technology developed for AI Strategist is far from obsolete. In fact, the core systems we built—particularly the document generation feature—are already being adapted for other projects.

For example, the real-time document generation system, which allowed users to receive a formal strategy document based on their AI interactions, can easily be repurposed for industries like legal services, where clients need detailed reports or contracts generated from conversations with an AI. Similarly, the RAG and embeddings work we did will be applied to projects where large datasets need to be queried quickly and effectively, such as in research-driven industries.

By repurposing these features, we’re ensuring that the time spent developing AI Strategist continues to add value to future projects, enabling us to offer tailored AI solutions to our clients in specialized markets.

What’s Next: Exploring Vertical AI Solutions

Looking forward, we are focusing on building vertical AI solutions that solve specific problems for defined industries. This strategic pivot allows us to create highly specialized tools that are less likely to be disrupted by general AI solutions from larger players. We’re already exploring new opportunities in industries such as healthcare, legal, and finance, where tailored AI applications can provide significant value by addressing industry-specific challenges.

The experience with AI Strategist has strengthened our commitment to developing products that focus on specific, high-value use cases, allowing us to leverage our technical expertise while creating defensible market positions.

Conclusion and Key Takeaways

The AI Strategist project may not have reached the market as originally planned, but it was a critical step in our growth and development at Fractal Labs. The lessons learned and the technical skills gained during the process have equipped us to navigate the rapidly changing AI landscape and apply these innovations to future projects.

Key Takeaways:

• Agility is essential: Pivoting quickly in response to new market realities can save resources and allow you to focus on more strategic opportunities.
• Focus on vertical solutions: Niche AI products that address specific industry needs are more defensible against competition from large, horizontal AI platforms.
• Leverage technical advancements: Technologies like retrieval-augmented generation (RAG) and embeddings can enhance AI systems by enabling better data retrieval and more intelligent responses.
• Non-attachment to projects: Being able to step back and re-evaluate a project without emotional attachment allows for better decision-making in a fast-moving industry.
• Repurpose technology: Even when a project is shelved, the systems and features developed can be adapted for other use cases, adding value to future projects.
• Prioritize real-time value: Features like document generation from AI interactions offer tangible outputs that can be reused across industries such as legal, healthcare, and finance.

‍

Looking for a reliable software development partner?

If you’d like to discuss a project, we'd love to chat with you! Let’s book a meeting and talk about your ideas, challenges, and how we can help you achieve your goals.

‍