INBOUND 24

Dharmesh Shah on The Future of AI Agents | INBOUND 24

6 min

When Dharmesh Shah took over the session, the presentation screen showed 17, where 1 was animated like a lit-up candle. Soon, the number 7 was swapped with 8, and it became 18.

On this note, he announced that HubSpot turned 18 this year!

Every year, at INBOUND, he loves to showcase the projects he has been working on for the whole year. In INBOUND 2024, Dharmesh brings Agent.AI—his 1500+ hours of hard work, along with different AI Agents.

Let’s see what Dharmesh, the Co-founder & CTO of HubSpot, has to discuss in the session, HubSpot Spotlight.

2024 is the Year of AI, AI Agents

Dharmesh recalled his INBOUND session in 2023 where they came up with ChatSpot.ai. This year, ChatSpot.ai became Breeze Copilot. This transformation was necessary for HubSpot to stay ahead of the rapidly upgrading AI trends.

AI is growing, and to understand it, Dharmesh introduces three Ds in the session. They are Debrief, Demystify, and Development.

Let’s understand them in detail.

Debrief

AI models are growing BIGGER!

Not just bigger, but FASTER; they are doubling every 6 months!

They are innovating in multiple dimensions. One of such highly-capable AI models is OpenAI’s newly launched GPT o1.

It has a surprisingly high level of reasoning capacity, like PhD level reasoning. To match with such AI tools, Dharmesh suggests that “we shouldn’t go to where the AI is; we should go to where the AI is going.”

He also mentions that modern AI models now take images, videos, audios, and text as inputs. For instance, instead of just code, these models can visualize a website like we see it.

To keep up with these quick advancements, HubSpot brings different AI Agents. They may look mysterious at first, so Dharmesh says let’s demystify them.

Demystify

Let’s first understand what an agent is and why having an agent is essential.

Dharmesh prefers to define an AI Agent the way Yamini described it in her session on AI Acceleration & Transformation:

“An AI Agent is a software that uses AI and tools to accomplish a goal that requires multiple steps.”

AI Agents are capable of solving all tasks, from the simplest to the most complex ones. Dharmesh adds an interactive slide here demonstrating the importance of every AI Agent:

AI Agent
Every AI agent is made up of three things:

AI Models + Tools + Memory to remember things across a task

An agent needs to remember past commands from the users till the whole task is completed. Or else, Dharmesh says on the funny note, “It’s like me walking up to the refrigerator, opening it and then forgetting why I opened the refrigerator.”

Now, is the time to see what Dharmesh and his team of experts have developed for streamlining these AI Agents.

Development

Dharmesh was over the moon to share his new product, but he made a disclosure first:

“The upcoming segment is rated X, X for experimental.”

Here, he introduces Agent.AI, an experimental but to-be successful project that he has worked on for over 1500 hours.

So, what is Agent.AI?

Dharmesh defines Agent.AI to be the number one professional network for AI agents. It’s also the only professional network for AI agents.

What? A professional network of AI Agents? Why do AI Agents even need a network of their own?

There are three predictions that Dharmesh makes for having an AI Agent Network:

Prediction 1: Future teams are going to be hybrid

In future, teams will have humans and AI agents working together to fulfill different tasks.

Prediction 2: Most of us will be using AI Agents this year

HubSpot users are already using AI Agents like Content Agents, Prospecting Agents, Social Agents, etc., to simplify and speed up their tasks.

Prediction 3: Many of us will be building AI Agents this year

For this, Agent.AI will provide a feature called Agent Builder. Even users with zero coding experience can build their AI Agents in just a few clicks.

Now, let’s understand Agent.AI in depth.

Agent.AI - Everything Dharmesh Discusses About It

On your browser, when you navigate to Agent.AI, and you’ll see the below given screen:
AI Agent
Every AI Agent listed in Agent.AI will have its profile, and from there, you can hire, follow, and add to team that agent.

So, basically Agent.AI is a collection of different AI Agents where new agents are added every week. Now, let’s see some common challenges that can be resolved with these agents.

Challenges that AI Agents Resolve

Here, Dharmesh states different challenges and how Agent.AI helps to overcome them:

Challenge 1: The number one challenge for marketers today is generating traffic and qualified leads.

Solution: The Conversion Rate Optimizer Agent. For instance, there’s a decent looking website, but it converts less. To optimize this site on the basis of what to improve in it and what to remove, will consume a lot of time.

Experts will have to review the content, check CTAs, tweak headlines, update testimonials, and whatnot. Meanwhile, this hectic task is made easier with the conversion rate optimizer agent.

You need to add your website link here and click on Go. With its expertise, this tool will generate a list of suggestions that you can implement to improve your site and its conversion rate.

AI Agent

Challenge 2: Sales reps spend the majority of their time researching prospect companies.

Solution: The Company Research Agent. Dharmesh recalls that Yamini mentioned salespeople spending the majority of their time in preparation and finding companies. This agent will save their time.

To find anything about the company, you’ll need to add that company’s name or domain here and click on Go.

In return, this agent generates a table of information of this company that consists of all the information like:

-> Overview, Research Questions Founders, Competing/Related Companies, Web Traffic, Twitter Follower Count, Team Growth, Organic and PPC Spend, Traffic, Keywords, Organic Keywords, and Tech Products.

You also get an option of Track Company to keep yourself updated with what new things are happening in that company.

Many other agents in Agent.AI can overcome different challenges, so sign up now and start using them.
AI Agent

Now, what if you come to know that an Agent can use another Agent to resolve challenges? Let’s see how!

Agents can Use Other Agents

After solving challenges for sales and marketing teams, Dharmesh now focuses on customer service.

For instance, a customer support rep will have a call with a customer, so before that, he wants a brief overview of their company and how it’s performing in HubSpot CRM and the outer world.

In this situation, they can use a Customer Briefing Agent. This agent will use a company research agent and other tools to provide insights into how that particular company is performing.

After reading about AI Agents, and what challenges they solve, you might be thinking, why is this HubSpot, a customer relationship management system, building Agent.AI? Let’s see why.

Why is HubSpot Building Agent.AI?

HubSpot is building Agent.AI to become a next-gen customer platform. As the AI trends are changing, and customer challenges are becoming complex, it’s getting essential to use new AI-powered features to keep up with market trends. These features will ultimately save time and resources of companies, and drive success to their business.

On the Ending Note

Both platforms provide significant eCommerce features to suit the varying needs of online stores. It includes a secure socket layer (SSL) certificate, the ability to sell products, and more. BigCommerce offers more built-in features, whereas Shopify provides a plethora of more than 8,000 free and paid third-party apps in its app store.

But the number of users has already surpassed the total of 47000!! And almost 1700 people are signed up on the waiting list to use the Agent Builder feature.

These numbers motivate him to publish the project soon in INBOUND 25, along with other new features that might help people improve their productivity.

That’s all from the INBOUND 24 series. We hope that you have gained all the insights on what new AI-powered features were introduced in HubSpot and how to leverage them to overcome different challenges.

For more updates, stay tuned with us. Happy reading!

Share:
Knowledge Base

Related Blogs

Dive into other interesting, well-researched, and nicely structured blog posts

Time For a CTA

Contact Us

Get a Quick Quote

We will strategize our execution based on your requirement