Article by Stacey Varsani
Founder & Principal Consultant, Hadouken Consulting
When I set up my business 10 years ago, I thought that my experience managing large P&Ls at a global company had prepared me for everything. I had overseen operations, sales, marketing, finance, HR, HSEQ, and contracts – but running my own business was an entirely different challenge.
Suddenly, I had to build systems from scratch, navigating contracts, proposals, pricing, payroll, financial statements, and more. It was a steep learning curve, and it took up a lot of time – time that could have been spent on client work or business development.
Using LLMs like ChatGPT has saved me a lot of time over the past year and would have helped me navigate many of the things mentioned above. However, the real game changer for many businesses, from start-ups to large global corporations, will be AI agents.
What Are AI Agents vs. “Normal” AI?
“Normal” AI typically refers to task-specific systems that excel at a single, narrowly defined function. For example, a customer service chatbot can answer frequently asked questions, or an AI-powered analytics tool can identify trends in sales data. These systems are powerful but fundamentally limited in scope.
AI agents, on the other hand, are designed to carry out multiple, interconnected tasks to achieve a goal with minimal human input. They are essentially virtual workers or decision-makers that can analyse information in an iterative manner, prioritize tasks, communicate with others, and refine actions based on previous outcomes. These abilities make AI agents more than tools – they are dynamic collaborators in problem-solving and decision-making.
AI agents have the potential to revolutionize how we work and innovate. While traditional AI systems are powerful assistants, they still require significant human intervention to define tasks and interpret results. AI agents, in contrast, offer the promise of higher-level automation. They can be given a goal, such as “increase website traffic by 20%” or “find the most cost-effective supplier,” and figure out the steps needed to achieve it.
For businesses, this means less time micromanaging technology and more time focusing on strategy. It’s like having a highly skilled team member who never gets tired, retains information from every interaction, and can operate 24/7. Taking this further, businesses can implement Multi-Agent Systems, where multiple AI agents communicate and exchange information with each other, allowing them to collaborate, compete and negotiate to achieve shared goals.
Use Cases for AI Agents
AI agents are quickly gaining traction. According to BCG, the market for agents is expected to grow at a 45% CAGR over the next five years. Deloitte estimates that 25% of companies that use generative AI will launch AI agent pilot projects or proofs of concept in 2025, growing to 50% by 2027.
Let’s look at just one agent that is currently on the market to illustrate the range of use cases across Sales & Marketing. Salesforce’s Agentforce can do all of the following:
Business development: The agent can autonomously answer product questions, handle objections, and book meetings for sales reps, using responses grounded in business data.
Pitching and negotiation: It can help salespeople practice pitching and negotiation with role-plays tailored to each deal. It then assesses its own ROI by analysing success rates of deals with coaching vs. deals without coaching.
Customer interaction: It can speak to customers on WhatsApp or websites to act as a personalized shopper.
Marketing: The agent can run campaigns end-to-end, generating campaign briefs, targeting audience segments, creating content, and analysing performance 24-7 against KPIs to recommend improvements.
Agents are being developed across a range of other functions and use cases. In logistics, Repsol just announced a partnership with Accenture to integrate AI agents to enhance logistics processes. These agents can handle a range of tasks, from predictive maintenance to real-time inventory management and supply chain management.
An example of an HR application is WebWork’s AI assistant, which can analyse employee time tracking data and provide actionable solutions in relation to productivity issues. It can even identify employees at risk of burnout to allow employers to take proactive measures.
The End of Surfing the Web?
Platforms like AgentGPT and Auto-GPT allow anyone to build or use pre-designed agents. Auto-GPT has one pre-built agent that can generate viral videos from topics trending on Reddit. Another can subscribe to a user’s YouTube channel, transcribe videos, identify the most impactful quotes, and generate social media posts to promote the videos.
Even browsing the web is changing. OpenAI’s Operator and Anthropic’s Claude can navigate websites, search, make purchases, and even book reservations autonomously. Google’s Mariner, with its “Deep Research” feature, can gather, analyse, and synthesize information into comprehensive reports, akin to a skilled human assistant.
While challenges remain – such as navigating complex web interfaces or avoiding infinite feedback loops – success rates have surged to nearly 60% over the past month (source: WebArena leaderboard). Soon, much of our online activity could be managed by AI agents, with humans stepping back to oversee and guide these virtual assistants.
Conclusion
AI agents are not just a tech trend – they represent a paradigm shift in how things get done. By combining autonomy with intelligence, they have the potential to enhance productivity, reduce costs, and unlock new levels of innovation. While it’s still early days, the possibilities for both business and personal applications are as exciting as they are vast.
Ready to harness the power of AI agents and virtual employees? Let us help you craft a strategy that drives growth and sustainable value creation. Contact us today to get started.
Comments