Deprecated: Using null as an array offset is deprecated, use an empty string instead in /home/u876752588/domains/capria.vc/public_html/wp-content/plugins/jet-engine/includes/components/blocks-views/dynamic-content/manager.php on line 113

TO: Capria Global Team and Advisors

January 6, 2025

Over the past two years, we’ve all witnessed the explosion of generative AI, transforming industries and creating immense opportunities for businesses that harness this technology effectively. As a venture capital firm, we focus on companies that use generative AI to enhance scalability, profitability, and operational efficiency—particularly at the application and services layers. As the media remind us daily, agentic AI is poised to become a cornerstone of business transformation. 2025 will be Capria’s “Year of Living Agenticly.” By embracing this challenge, we can enhance our own operations, deliver unmatched value to our portfolio companies, and solidify our position as applied AI leaders across the Global South.

This memo outlines our strategy to deepen our expertise in the agentic AI wave, positioning ourselves as better investors and enablers for our portfolio companies to thrive in this rapidly evolving landscape. There is no doubt that for us to succeed as the top decile investors we aspire to be, we must go even deeper and faster in how we learn, experiment with, and deploy AI for ourselves as well as our portfolio.

What Is Agentic AI?

Many of you have already read about and even experimented with agentic AI. For those who have not, this HBR article is very helpful in understanding the broad context of agentic AI. Here’s a synopsis: agentic AI refers to artificial intelligence systems capable of autonomously performing complex tasks with minimal human oversight. Unlike traditional AI, which often relies on predefined instructions, agentic AI integrates decision-making, adaptability, and execution to achieve goals in dynamic environments. These systems use generative AI capabilities to independently process information, learn from interactions, and execute actions, making them ideal for use cases like customer support, workflow automation, and personalized recommendations. Key attributes of agentic AI include:

• Autonomy: Operates with minimal human intervention, making decisions and executing tasks independently.
• Goal-Oriented Behavior: Focuses on achieving specific outcomes, such as optimizing business processes or enhancing user experiences.
• Adaptability: Learns and evolves based on new data and real-time interactions, improving performance over time.
• Low-Code/No-Code Accessibility: Often integrated into platforms that allow non-technical users to deploy and customize AI-driven solutions efficiently.

The Momentum Behind Agentic AI

Agentic AI is gaining significant traction. Investment in AI startups reached $25 billion globally in 2023 (CB Insights). Leading incumbents, such as Salesforce, have accelerated their efforts in AI, introducing platforms like Einstein GPT, which integrates generative AI into customer relationship management systems, followed 6 months later with their announcement of AgentForce where CEO Benioff outlined that they are all-in on agents. The list goes on for a long time…

As reported recently in AI Weekly, enterprises are embracing agentic solutions at a surprisingly fast pace. “In a recent survey of 1,100 tech executives conducted by Capgemini, 82% of the respondents said they intend to integrate AI-based agents across their stacks within the next 3 years — up from a current 10%. More importantly, as many as 70 to 75% of the respondents said they would trust an AI agent to analyze and synthesize data on their behalf”.

VCs are also doubling down on low-code and no-code agentic AI platforms that democratize access to advanced AI functionalities. These platforms empower non-technical teams to deploy AI solutions efficiently, driving adoption across diverse sectors. According to Gartner, by 2025, over 70% of new AI-driven applications developed in enterprises will use low-code/no-code platforms. The shift to Agentic AI will accelerate this further.

Note that despite the momentum and enthusiasm, there are many structural issues to be addressed before AI is a profitable and environmentally sustainable addition to the global economy. And there are security issues too – OpenAI is reportedly holding back on their agent release due to concerns around prompt injection. For a more context and a curmudgeonly view, read Gary Marcus “25 Predictions for 2025”.

Our Experience with Generative AI

In 2023 and 2024, we explored generative AI through chatbot technology built on Retrieval Augmented Generation (RAG) architecture, with a light agentic layer on top for certain implementations. These efforts helped us:

• Learn the nuances of generative AI implementations
• Improve internal understanding of the tech
• Successfully guide portfolio companies on adopting and productizing the MVP implementations we create for them and also to build, integrate and scale their own tech stack

However, as the field advances, staying ahead requires us to expand our focus to include deeper agentic AI. We’ve started delivering some light-weight agentic features via GenericBot to a few of our portfolio companies – this is a good start. Now’s the time to go to the next level. This shift is crucial to identifying how our portfolio as well as new pipeline startups can benefit from transformative low-code or no-code agentic platforms, validating their effectiveness, and increasing value.

One of our portfolio companies, Agrofy, has already deployed an agentic RAG framework for their pre- and post-sales chatbot, named Clemen. The addition of the agentic tech to the front end allows the chatbot to draw from multiple company data sources based on the need identified in the customer’s query and to use the LLM to assemble a cogent reply based on the information gathered.

Will AI Agents Replace Any of Us Anytime Soon?

Given the demonstrated pace of adoption coupled with the future-looking promises (sometimes outlandish) made by masterful tech hype-masters, it’s hard not to wonder if AI will be replacing any or all of our jobs. Many of you know I’m writing a book, Flourishing with AI, to address exactly this concern. My views have not changed over the past 6 months: there are many things current and near-term (2-3 years out) AI tech can do to make our jobs easier and more efficient, but there is no job at Capria that could be even 50% replaced by AI in the near term. Why? First, LLMs on their own suck at working with numbers, as we live and die by numbers. New agentic models will make big improvements in this area but it will take time for them to permeate the complexity of boutique financial companies like ours. Even trusting a financially literate agentic LLM to do anything on its own with numbers without deep oversight by us is not happening anytime soon. Second, while we all have many things that AI already does and will continue to do more of to make our jobs easier, there are so many people-oriented aspects of our work that are far from the reach of AI. All that said, AI can do more for each of us which will free up time to do more strategic and higher-level tasks that AI cannot do. This is critical for us as we will not be adding many new team positions even as our portfolio grows. Also, we will have a considerable increase in our (exciting) efforts to realize strong exits for many of our portfolio; this is still a new muscle we are building. To deliver on this, we will all need to become more productive — much of which will come with applying GenAI tools. Getting going on increasing our own productivity is a big part of this call to action.

How Will We Address this Call to Action?

Our GenAI team will take the lead on tech and others have also been assigned specific areas to dig into. We’ll do more planning before our March offsite in Istanbul and look to go deep in a few areas that week as well.

Conclusion

Chatbots were the 2023/24 drivers of AI adoption, led primarily by nearly one billion consumers using them as of the end of 2024. Agentic AI is expected to be the next wave and likely will be the wave of innovation that drives much more corporate adoption of AI. By ensuring we are leading edge adopters of the tech for our own use and that we continue to support our portfolio companies to adopt ahead of others, we will ensure a strong position as the leading Global South Applied AI investors and will deliver commensurately leading returns to our investors.