Apply to Register Here

Founded in 2015, Activant is a global investment firm with $1.5 B AUM that partners with high-growth founders and teams that are transforming commerce.

Activant

We’re always looking to learn about new perspectives, new research, and new ideas.

Get in touch

Disclaimers & Terms of Use

Privacy Policy

323 Railroad Avenue
Greenwich, CT 06830
USA

Greenwich

New York

Berlin

Constantia Uitsig Wine Estate
Spaanschemat River Road
Cape Town 7806
South Africa

Cape Town

Harish is providing the Amazon Prime fulfillment experience to everyone else through a combination of AI and an asset-light network of warehouses across the United States.

Alex is automating workflows for the world’s largest enterprises by using process mining to build the connective tissue for businesses.

Mayan is delivering Autonomous IT to the largest companies in the world through systems of intelligence.

Carsten built one of the largest commerce platforms in the world, which powers enterprise merchants and retailers across the globe and now serves as the SAP commerce engine.

Soups is stopping financial crime in its tracks by building the first end-to-end platform for fraud, risk, and compliance.

Adam is enabling restaurant owners to take back their customer relationships by digitizing their entire online and offline operations in a single platform.

Activant is a research-led growth firm that invests in relentless founders

Read past issues

Monthly Newsletter

Browse Research Library

Research

Perspectives

Home

Activant is a research-led global investment firm that partners with high-growth companies.

We partner with founders who have won their initial battles and are ready for the next challenge.

We typically invest when two things come together. First, the company has the potential to achieve escape velocity and transform an entire industry. Second, we believe the founders can lead the company through hyper growth, and our support can significantly increase the probability of this success. 

The critical time window to escape competitors and transform an industry is not open for long.

That’s why after we invest, the real work begins.

About

We only partner with a few founders per year. But when we do, we’re fully committed.

Our portfolio

Companies

Read our investment rationales and market perspectives.

A team built to support and advise yours.

We invest behind insights.

View all research

Founders who aspire to build something bigger than themselves chose to partner with us at their most critical point.

Team

A born and raised Texan, Julius holds a passion for ecosystem-changing technology. Having worked as both a software engineer and an investor, he aims to provide tangible technical value to operators that he feels is sorely lacking in the overall industry. All too often, the narrative outstrips the technical conversation, and by understanding what it is founders are building, Julius believes that he can provide increased leverage to them as they grow. 

In his free time, Julius keeps up to date on the latest NLP research and algorithmic trading strategies.

"I joined Activant to help add technical expertise in a traditionally finance-dominated world. Activant has this incredible ability to bring people from different fields together and working to invest in truly disruptive companies. For me personally, I can’t wait to work with founders shaping the future."

Trucking is a vital cog in the U.S. economy and a way of life for more than a million over-the-road (OTR) truckers and three million commercial fleet 

 Together these groups spend a combined $200+ billion on diesel and gasoline every year, and another $150+ billion annually on tolls, maintenance, lumpers, taxes, and a host of ancillary 

 For the privilege and convenience of using a fuel card, the issuers take a cut of 1%-2%. Sound like a good deal? Yes, for the card issuer, but not for the trucker.

Once the rubber hits the road, truckers quickly find that the great “in-network” deal they were sold on doesn’t hold up to the pump price across the street. They’re told about extensive fuel card networks, then have to drive an extra half hour to find an in-network station, paying the fuel cost of going out of their way as well as the lost-time cost. Finally at the pump, drivers find themselves waiting in long lines, wasting even more time while the clock continues to tick.

That’s just the on-the-road experience. Add in a host of account maintenance, transaction, and card replacement charges, and all-in fees are more than likely 50% higher than what truck drivers thought they signed up 

 And because Fleetcor and Wex – the two dominant fuel card companies – control nearly 70% of the $375 billion gross spending OTR card market, they have little incentive to innovate or improve their technologies and product offering. The closed-loop systems they operate present a litany of unforeseen expenses.

However, there are some innovative start-ups targeting this $7.5 billion market that is ripe for 

TAM Analysis

Fuel cards emerged in the 1980s to solve the cash flow and tracking challenges that truckers and trucking companies endured. Early leaders like Comdata and EFS were ultimately surpassed by Wex and Fleetcor, which began acquiring and aggregating more than 150 companies to largely consolidate the market. Today, 99% of all truck stops offer payments through either Fleetcor or Wex, and most fuel card agreements in OTR flow through their cards.

Drivers using a plastic charge card issued by Fleetcor or Wex must go through a series of prompts while inputting multiple codes, allowing the card to attribute the transaction to the driver and the specific trip. Importantly, after the transaction is completed, the fleet pays Fleetcor and Wex (i.e. the card issuer) for the fuel, and 

Fleetcor and Wex pay the truck stop, acting as the merchant acquirer. This creates a closed-loop system, in which Fleetcor and Wex act as the card issuer, the transaction processor and the merchant acquirer.

For the truck driver, this has a few notable benefits. First, because the drivers pay Fleetcor and Wex directly for the cost of the fuel, and the card-issuers then pay the truck stops, the card issuers can afford to give drivers cash prices while providing the flexibility and convenience of a plastic charge card. That represents 2-3% in savings per fill-up. When the average fill-up for OTR trucks is 

, or about $500 to $600, the savings can be $10-$20 per fill at today’s 

 Secondly, because Fleetcor and Wex capture both sides of the transaction, they have near-perfect data visibility, which can help to combat fraud. Closed-loop systems can see what type of fuel a driver purchased, how much they paid, and what trip they’re on. If the driver happens to be in Cedar Rapids, Iowa, and a fraudster has used the same card details to buy $1,000 of fuel near Cedar Springs, Texas, they can easily dispute that charge using transaction metadata contained within the system. These are proven positives of closed-loop systems.

Recent innovations in credit card technologies and metadata tracking have, however, made it easier and cheaper to automate expense tracking and fraud protection, making one of the major benefits of fuel cards redundant. Closed-loop systems also have significant downsides for truck-drivers and for the truck stops in their networks.

To understand the drag fuel cards place on the industry, it’s helpful to look at the end-to-end process and highlight the many inefficiencies and unnecessary costs of filling up a vehicle.

Before a driver can fill their vehicle, they need to use the Fleetcor or Wex app to figure out which truck stops are in their fuel card network. Conversations with industry leaders suggest that drivers often need to travel up to 30 minutes off their path to fill up at an “in-network” fuel station. This costs them twice: first, in the extra gallons spent driving to the truck stop and, secondly, in the time eaten out of their 11-hour daily limit.

Truckers often find that the price posted at a “Big 3” stop (Love’s, TA, Pilot Flying J) is meaningfully higher than at small, regional chains. For large fleets, which negotiate directly with the truck stops and operate on a cost-plus model, this isn’t really an issue. But for the smaller fleets and owner-operators, who get price-minus costing, this means a higher “discounted” price for in-network fuel than they could get across the street at a regional chain.

Besides potential waiting time at the truck stop island, the average driver takes between five and 15 minutes to enter the long list of numbers and selections required to identify their trip: mobile app codes, reward card details, unit numbers, trip numbers, fuel types, DEF selections, and more. And while the actual pumping of fuel takes less than four minutes, if the driver gets any number wrong, they will have to restart the whole process.

All of this time is on the clock, adding to the driver’s 11-hour maximum behind the wheel. When drivers are paid by the mile, this translates directly to lost earnings.

Away from the pump, a driver might be hit with a late fee, a card recharge fee, a toll fee or even an unused money code fee. On average, we found that smaller carriers and owner-operators pay up to 50% more in fees than they expected. In any given month, this can add up to thousands away from a small carrier’s bottom line.

Card theft or skimming is always a risk. Fraudulent charges are common and drivers reporting unusual transactions may have their cards frozen for up to 60 days as Fleetcor and Wex investigate before sending a new card, leaving the driver back at square one using cash. Also, because Fleetcor and Wex are struggling to push chip cards, most drivers are stuck with old technology magstripe cards, years after their personal credit cards got the more secure chip.

Many of these pain points are a feature — not a bug — endemic to the current fuel card duopoly. In addressing these challenges, Activant research highlights the opportunity for open-loop alternatives.

New solutions are emerging to address the pain points drivers face.

Over the past few years, both Visa and Mastercard have rolled out better data requirements for fuel merchants wanting to accept their cards. Now, 98% of truck stops offer Level 3 data to Visa and Mastercard, meaning the price per gallon, the gallons pumped, the DEF and refrigerant selections can all be logged in one open-loop 

 The result is that the long-time expense management benefits of having a closed-loop network are disappearing, as open-loop cards can offer the same visibility without the network lock-in.

Electronic Logging Devices (ELDs) come standard in all new OTR trucks (by law) and in most commercial vehicles used in fleets. These devices can be connected to a cellular signal and can provide real-time data about the GPS location, mileage, fuel level and MPG of the vehicle remotely. The natural next step is to integrate this data into the fuel payment workflow, which can immediately eliminate buddy fraud by allowing only the necessary gallons and stop theft-driven fraud by verifying where the truck is filling up.

Newer fuel cards can provide a modern workflow that eliminates the long codes a driver currently punches into the keypad when they fill. The industry has already seen some early versions of this, with apps that let drivers make selections with a tablet instead of the keypad, but this has further potential. Imagine a world in which your integrated ELD system already knows how much fuel a truck needs, what load it’s carrying, and whether the driver needs to fill up DEF. This future is being built today by some of the players mentioned below.

As data visibility increases, Activant anticipates a cardless world in which RFID tagging and GPS geo-locating identifies the vehicle at the pump, knows the number of gallons to pump from the ELD system, and accepts ACH direct-debit (cash-priced) payments from a known account without the need to swipe a card. To realize this vision, core infrastructure at truck stops and gas stations will need to be reworked to accommodate the technology, but it will ultimately have to happen.

When we studied the market, we chose to segment the players based on the stakeholders served and where on the spectrum from a point solution to a full platform offer the product fit.

Trucking Payments

Three categories of start-ups emerged, each tackling the pain points in unique ways.

These are open-loop cards, targeting owner-operators and small-to-midsize fleets (~1-100 vehicles). They offer universal acceptance, modest discounts, lower fees, and better product user interface and flexibility. We believe they offer a differentiated value proposition, both from a product and a cost perspective, and they have seen solid traction (to the tune of tens of thousands of cardholders). It’s worth noting that most of these challengers are powered by Visa, which is committed to growing their fuel card payments network, to catch not just Fleetcor and Wex, but also to compete with similar efforts from Mastercard.

, which is mainly focused on mid-market OTR carriers, 

, which targets non-trucking commercial fleets, and 

Another category of challengers are logistics tech companies that are starting to offer fuel cards. Usually, these companies already have an existing adjacent solution targeted at carriers and are looking to expand their value proposition.

Emerging winners in this segment include companies like 

, which started out handling lumper, warehouse, repair and accessorial payments, all of which are much simpler than fuel. Relay just launched its own fuel card and Roadsync could follow suit. Other companies, such as 

, which offer a broader trucker workflow solution (often including load boards), as well as discount fuel networks such as 

, are also offering fuel cards. Brokerages like 

 have launched fuel solutions with existing partners like AtoB. Even 

, one of the ELD leaders, has started offering their own cards. New fintech issuer processors such as 

 have led the way on powering logistics payment solutions.

There are challengers looking to fundamentally reinvent the experience, offering a platform solution comprised of various combinations of hardware and software

 has built its own fuel payment network and is actively signing on carriers and truck stops in a very cost-effective manner. Similarly, 

 links commercial and personal cars to its own digital payment method, allowing drivers to pay for fuel without the need for a card.

If you’re an owner-operator or part of a fleet, you should seriously consider switching your card. You’re paying too much in hidden fees, higher prices, lost time, extra miles, and unnecessary fraud exposure.

We’re excited about the future of fuel cards and what it means for the industry. There hasn’t been enough innovation in the last 30 years, but things are changing – one card at a time – and we have no doubt that open-loop winners will emerge in the near-term to challenge the entrenched (and very profitable) duopoly. If you're building in this space, 

The information contained herein is provided for informational purposes only and should not be construed as investment advice. The opinions, views, forecasts, performance, estimates, etc. expressed herein are subject to change without notice. Certain statements contained herein reflect the subjective views and opinions of Activant. Past performance is not indicative of future results. No representation is made that any investment will or is likely to achieve its objectives. All investments involve risk and may result in loss. This newsletter does not constitute an offer to sell or a solicitation of an offer to buy any security. Activant does not provide tax or legal advice and you are encouraged to seek the advice of a tax or legal professional regarding your individual circumstances.

Opportunities to innovate with fuel cards

Pain at the Pump

Application Programming Interfaces (APIs), the digital conduits that allow different software systems to “talk” to one another, are the unsung heroes of digital commerce in a world where the number of connections between entities is increasing exponentially. 

Businesses today use nine times more individual software tools than they did five years ago

  and complexity is only increasing. Maintaining these myriad integrations and the constant upgrades to each is a major challenge, particularly if engineering teams need to deal with each connection individually. Differences in time formats, currencies, accounting or reporting nomenclature, and a host of other nuances can make managing APIs at scale a headache.

There are solutions to this, and markets thrive when aggregators enable simple connections, much like 

 did with its universal API for fintech companies to connect to user 

” is an abstraction layer that handles communication with many different APIs and backend data models. Companies only need to build one integration – to the unified API provider — and it then manages the complexities of multiple API connections. Exhibit 1 illustrates the simple difference that a unified API offers for developers.

: changes to APIs can lead to downtime if the individual connections are not aligned, resulting in breakages

: fixing these breakages is often an inefficient process consuming expensive developer capacity

Accessing new integrations and data points

: access is possible to all the API integrations a Unified API solution has built

: Unified APIs offer significant time and resource savings, allowing more rapid prototyping and new product launches.

The case for Unified APIs is compelling and we expect the market to expand over time as API-first businesses proliferate.

We estimate that companies spend more than $50 billion annually in the U.S. alone on maintaining external-facing APIs. Our estimates include (1) the number of companies that use external API integrations, (2) the number of API integrations per company, and (3) the average annual maintenance spend per API. Based on our interactions with industry experts and practitioners, we conservatively estimate that companies have an average of four external API integrations, and that maintaining these costs approximately $620 per API per annum. Exhibit 2 summarizes the results of our market sizing analysis.

Victory Matibiri, CEO and Co-Founder, Mansa

Estimated Unified APIs Market Size

In our market study, we differentiated the players based on the time it took to integrate their solutions and the degree of customization of data fields possible. The Activant Market Map is shown below.

One of the benefits of the segmentation approach used is the alignment with the broader market of solutions addressing the pain points of building, maintaining and orchestrating APIs. Each quadrant in the market map represents a unique approach to API integration.

 do what the name suggests: they unify APIs. These tools are generally faster to integrate as they only need to process a single set of API documentation. Unified APIs apply a data normalization technique that standardizes API fields across software vendors. The more coverage a business has across vendors, the more participation it can expect.

 solutions help internal development teams build individual API connections in a low-code fashion. Relative to Unified APIs, these tools generally require more engineering hours to build but are ideal for scenarios where a high degree of customization in the data being pulled is required. These tools do not apply a data normalization technique.

Integration Marketplaces as a Service (iMaaS) 

solutions focus on the distribution of integrations and enable SaaS companies to provide APIs to their customers in the form of a marketplace. In the same way, a traditional marketplace helps organize and showcase products a business can sell, an integration marketplace is an effective platform to showcase all the apps an SaaS can integrate with.

uses company resources to build and maintain APIs. The resources required ultimately depend on the complexity of the integration, which includes factors such as the number of end-points to connect, the amount of data being pulled, and security requirements. Costs can vary from $10-$100K+ per integration.

Importantly, many of the companies that we have located on the market map offer iPaaS, iMaaS, and Unified API solutions. It is not a one-size-fits-all world and customers have the luxury of choosing a combination of the solutions depending on their specific needs. Simple solutions will win. Less friction is likely to result in more commerce.

It’s still early days, despite the clear need for a ‘middleware’ layer to help manage the complexities of multiple API integrations. We believe that tomorrow’s winners will be those that offer solutions with 

 to allow teams to pull customized data listings. Additionally, we think the use of Large Language Models (LLMs) to automate the construction of API integrations into their common data layers will become table stakes. This may decrease the cost per integration and has the potential to expand industry margins – which is likely to increase industry competition. Whatever the future holds, there’s definitely change coming to the status quo.

Activant’s conviction in the bright future for unified API solutions is summarized in our thesis map below.

, a first mover in the use of LLMs. The company recently unveiled 

 to provide LLM-built API integrations to customers.

, a solution powering the credit decision engines of top tier banks and FinTech companies that integrates with 19 accounting solutions and 16 commerce platforms. Lenders can quickly assess, for example, the size and quality of incoming payments to their customers by comparing contract details against invoices in accounting platforms.

, a market leader offering integration with 200+ Human Resource Information Systems (HRIS). Providers of 401(k) plans, HR technology, and fintech solutions can integrate with an employer’s HRIS system to streamline employee onboarding, track payrolls, update benefit deductions, and manage other core HR and compliance functions

, which offers its customers broad and deep solutions across disparate use cases in CRM, accounting, logistics, and the like, allowing them the option to build fast, then deep. A unified API allows customers to instantly set up real-time read-and-write integrations, and when it’s time to build more complex processes, customers can leverage Alloy’s Embedded iPaaS.

, a first mover in building the category across Africa, which allows fintech firms to read and write across 17+ African commerce, banking, accounting, and payments platforms.

At Activant, we’re excited to watch this market evolve. If you’re building in the space, have a different perspective, or enjoyed the piece, 

This content may not under any circumstances be relied upon when making a decision to invest in any fund or investment, including those managed by Activant. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Activant. While taken from sources believed to be reliable, Activant has not independently verified such information and makes no representations about the current or enduring accuracy of the information or its appropriateness for a given situation.

Activant does not solicit or make its services available to the public. The content provided herein may include information regarding past and/or present portfolio companies or investments managed by Activant, its affiliates and/or personnel. References to specific companies are for illustrative purposes only and do not necessarily reflect Activant investments. It should not be assumed that investments made in the future will have similar characteristics. Please see “full list of investments” at https://activantcapital.com/companies/ for a full list of investments. Any portfolio companies discussed herein should not be assumed to have been profitable. Certain information herein constitutes “forward-looking statements.” All forward-looking statements represent only the intent and belief of Activant as of the date such statements were made. None of Activant or any of its affiliates (i) assumes any responsibility for the accuracy and completeness of any forward-looking statements or (ii) undertakes any obligation to disseminate any updates or revisions to any forward-looking statement contained herein to reflect any change in their expectation with regard thereto or any change in events, conditions or circumstances on which any such statement is based. Due to various risks and uncertainties, actual events or results may differ materially from those reflected or contemplated in such forward-looking statements.

Building connections

Unified APIs

We are hardly the first to claim a new technology is poised to upend work as we know it. With every wave of innovation comes a chorus of such pronouncements, many of which turn out to be hyperbole. Technology may make our jobs easier, but most of the time, the core of what we do remains fundamentally the same. 

But spend five minutes on ChatGPT, or any comparable Large Language Model (LLM), and it’s clear: it’s different.

The reason is simple. Most technologies help us accomplish tasks. They’re tools, a means of making us more productive. AI, on the other hand, can think (or at least, provide the appearance of thinking). It’s like having another knowledge worker on the payroll.

Activant Perspective 1: LLM-enabled software will autonomously accomplish highly complex tasks.

This is the beauty (and threat) of LLMs: they can do work we thought would long be the purview of humans.

Given this, one could be forgiven for worrying their job will soon be automated away by an AI. In the short term, this will lead to dislocation. 

But technological revolutions are never zero-sum – in fact, they’ve always been accretive.

 As workers adapt, they become more productive, and the companies they work for can do more with less. Some stop there, to be sure, but most businesses opt to invest the gains back into their infrastructure and people. In the end, revenues go up, and the economy grows.

That said, we’re still in the very first days of this cycle – and critically, of the technology itself.

Activant Perspective 2: LLMs alone are not enough to achieve Perspective 1.

Case in point: a large language model can’t complete most tasks on its own. After all, an LLM is merely a dense set of attention networks; essentially, statistics on a massive scale. Yes, they can engage in human-like “thought,” but to generate real value from any of these models, a human would have to integrate them into actual workflows – and make use of their decision-making prowess.

For example, imagine that you hired someone to lay bricks. You could lead them to a stack of bricks and explain where and how you want them laid. They could perfectly understand your instructions. Yet, understanding isn’t enough – the worker needs to be able to, well, lay bricks.

In this case, they need to have more than just cognition – they need to have arms, legs, and physical strength. Furthermore, they need a memory to recall how you wanted each of the bricks to be arranged, and a way of correcting any errors so they don’t cause the eventual structure to fail.

Put another way, if LLMs are to extend beyond mere Q&A, they need (1) a suite of Tools to actually impact the real world, (2) memory repositories to remember what actions they’ve taken, and (3) auto-critique algorithms to error correct along the way.

When all these components come together, an “Agent” is born.

 An Agent is orchestration software that combines an LLM with memory, tools, and auto-critique algorithms.

 An Agent is a highly advanced bot that can complete work that has been assigned to it in natural language.

Though still in their infancy, Agents are poised to fundamentally reshape the way we work. They will become natural extensions of your organization and complete day-to-day tasks – akin to hiring more employees. They’ll help you get work done, even collaborate with others on your behalf. Eventually, they’ll be able to work together, breaking down tasks into atoms of work able to be handled by a full “office” of Agents, some of which orchestrate, while others specialize.

The result is why this time is different: working together, Agents will be able to handle increasingly complex tasks, replacing knowledge workers in a variety of fields.

 As they do, what it means to work – and to lead – will change drastically, as both will increasingly depend on a mix of human and human-like capital.

Activant Perspective 3: Agent Teaming is the future of autonomously completed, complex work – 

Few have sketched a path from where we are today, to the autonomous knowledge workers of the future. But that future is closer than you’d think, and one for which you almost certainly want a roadmap. To learn more about how this transformation will unfold, and how you can best position yourself within it, read on.

The Agent Ecosystem {2}

The Agent Ecosystem of the future has four key components:

provide “conscious thought” within AI Agents and have been trained on text data,

external APIs to facilitate task completion and Agent interaction with the real world,

 includes various data storage mechanisms to retain and recall information, and

the ability to correct mistakes when completing tasks.

To better understand the components, we imagine how each one might erode a job we know quite well: that of an analyst at Activant (sorry, John). Among many other things, we often rely on our analysts to create market maps, which requires them to:

Find companies that exist in that industry,

Compare these companies qualitatively, and

Place them on a 2-dimensional grid according to this comparison.

Until recently, this was almost inconceivable as a task for machines. Now, let’s explore how we’d put our “Agent Analyst” to use.

AI Agent Ecosystem

1. Understanding the Industry (and the Task) – Large Language Models

For our Agent Analyst to understand the task at hand, it requires the capacity to comprehend instructions stated in natural language. While LLMs provide that ability, they come in many different shapes and sizes (literally). For that reason, here we outline a basic framework for evaluating LLM providers, including both foundational model providers and companies offering Fine-Tuning as-a-Service (FTaaS

When comparing foundational models, the number of parameters in a model is often a good approximation of output quality. As a rule of thumb, the more parameters a foundational model has, the better its performance across different benchmarks, 

such as reasoning, web browsing, or even game playing

. Although fine-tuning a model can boost its performance on a particular task, larger models with more parameters tend to have better generalization skills – making them more useful.

As seen below, model sizes are increasing on an exponential basis with base 2, and although not pictured, model performance is similarly increasing – just not at the same rate.

Foundational Models Size Map

At the heart of any LLM lies its architecture: a blueprint dictating how various attention heads and layers are organized. Then, there are the "weights" — numeric values assigned to parameters in the neural network, refined and adjusted through training to enhance performance. Training involves feeding data into the model, allowing it to learn and make predictions, and then adjusting the weights based on its accuracy. It's through this process, often requiring substantial computational resources, that the model obtains its ability to understand and generate human-like text.

, “openly” share their architecture, weights, and sometimes even their training methodologies. This approach not only reveals the intricacies of their design but also democratizes access, allowing enterprises, researchers, and even enthusiasts to build upon, refine, or entirely repurpose these models for their own applications.

Conversely, some entities opt for a more guarded approach. Closed-source models, such as GPT-4 by OpenAI, often protect the details of their architecture, weights, or training processes, presenting them as somewhat of a black box. Instead of allowing direct model access, they employ APIs as intermediaries, thereby controlling and often monetizing the interaction between the end user and the model.

Foundational Models

“Provider Modality” relates to the data type a provider offers models for. This includes various transformations such as text-to-video, text-to-text, text-to-image, and image-to-text, among others. Companies that operate in just one domain are labeled as “Uni-Modal”; by contrast, “Multi-Modal” companies provide models spanning multiple domains. As Agents continue to evolve, we expect a growing emphasis on modality in LLM models, as those able to perform more diverse tasks and offer easier integrations will have a distinct advantage in the marketplace.

While Agents will no doubt depend on the 

, we actually expect most Agents to remain agnostic to the specific LLM they use. 

Today, OpenAI, Google, and Anthropic dominate the industry, but this is hardly set in stone. If anything, the next few years will see considerable competition, with whoever invests the most in R&D in a constant arms race. Given this uncertainty – and the promise of better models always around the bend – it’s hard to imagine Agents relying on a single provider. From an investment perspective, backing one LLM provider is essentially a vote of confidence in its R&D team, but past performance is not (necessarily) indicative of future results.

Having understood the industry (and the task), the Agent Analyst now needs the ability to search for companies. In other words, the Agent Analyst needs a Tool.

Tools are snippets of code that foundational models can use to create, modify, and utilize external resources in order to execute tasks they couldn’t otherwise complete on their own.

Most Tools are problem specific. In this case, all the Agent Analyst needs is a connection to Google, or SerpAPI, a prominent wrapper for Google’s Search API. Unsurprisingly, other tasks will require different Tools. Game this out, and you quickly arrive at the need for Toolkits and Tool Aggregators. As tasks become more complex, Agents will rely on not one, but a variety of Tools to accomplish a single objective. Through Aggregators, Agents will be able to access the resources they need, programmatically, and as those needs arise.

AI & LLM Tooling Companies

Single Tools refer to companies or services that specialize in just one function, such as searching the internet, booking a flight, or creating realistic voice audio files. While these offerings do have certain strengths, integrating them into a Toolbox, and thus into Agents, is often cumbersome. Notably, in prominent Agent libraries

, Toolboxes need to be set up before an Agent itself is even initialized. 

This means if there's a need to integrate additional Tools later – for example, if an Agent encounters a new task that requires a Tool outside of its preset toolkit – the Agent would have to initialized all over again.

 Given these considerations, embedding Single Tools into an Agent's workflow can be quite labor-intensive and time-consuming.

AI Agent Tooling

By contrast, Tool Aggregators offer Toolboxes-as-a-Service (TaaS). Theoretically, aggregators can compile an infinite number of Tools and distribute them via a single, unified API. In practice, they typically provide Tools that have been identified by an Agent as required to complete a given task. If an Agent requires determines it needs additional tools, however, it can access them without restarting.

 companies could provision Tools and/or provide tool-building functionality for Agents. As long as they offer an API, Agents can leverage these solutions as they do Tool Aggregators, integrating (and even custom-building) Tools into their workflows on the fly.

Many LCNC platforms have existed prior to the advent of LLMs and already have large libraries of powerful, ready-built Tools to draw upon. Several companies, including 

, have begun offering Agents-as-a-Service (AaaS), using these libraries to jump ahead of competitors, many of which are starting from scratch. Others may choose to offer APIs to facilitate quick adoption within the Agent Ecosystem and more reliably maintain their market share.

Because of their ease of integration and use, we believe Tool Aggregators and API-based LCNC platforms are the best positioned to take advantage of this new Agent ecosystem

. As platforms like LlamaIndex and Relevance AI grow, Agents will be able to dynamically access a vast library of toolsets without needing to update their structure or code – increasing both their functionality and value.

As our Agent Analyst has progressed, it has gained the capability to discern tasks and search for companies. However, identifying a company means nothing if an Agent can’t remember it for later use. To facilitate this, Agents require a dedicated memory component that can store companies and their associated descriptions.

Memory

Over the years, the software industry has seen a multitude of data-structure innovations, from traditional SQL-based relational tables to modern streaming services like Kafka. Yet, to truly grasp the intricacies of Agent development and the depth of virtual cognition, it's helpful to draw parallels with human memory.

At a high level, human memory is split into three core categories: sensory memory, short-term (or working) memory, and long-term memory.

In humans, sensory memory can be thought of as the direct inputs from our various senses — taste, smell, sight, touch, and hearing. These are perceived subconsciously, processed then filtered for relevance. Similarly, in the realm of Agents, 

A Technical Guide for Founders & Operators Building Agents

Towards the AI Agent Ecosystem

Activant’s resident technical expert. Brings his engineering background to finding and partnering with founders and their teams.

Office

Hometown

Residence

Julius
Stener

Investment Team

Office

Hometown

Residence

Research

Office

Hometown

Residence

Julius Stener

Investment Team

Office

Hometown

Residence

Research

Julius
Stener