Episode 106: How Autonomous Agents Drive Business Growth at Scale with Dikshant Dave | EP106

Clearly we do not think that AI agents have to replace humans and the good thing is that most businesses that we talk to are not looking at that either.

Most businesses are saying how can we do more of what we are doing, so at least that's a good sign that even the leaders are saying I have a team of 40 people and this is our throughput.

The first thing is can we increase the throughput because I'm not interested in necessarily cutting the cost out; I want to grow the business.

Hi, my name is Demetri Panici and I'm a content creator, agency owner, and AI enthusiast.

You're listening to the AI agents podcast brought to you by Jotform and featuring our CEO and founder, Aytekin Tank. This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show.

Hello and welcome back to another episode of the AI agents podcast. In this episode we have the founder and CEO of Ziggman, Digshant Dave. How are you doing today, Dan?

Very well, Demetri. Thank you for having me here. How are you doing?

Absolutely. I'm doing well and very excited to chat with you about Ziggman. It seems like I like the tagline on the top of your website. It says engage every customer and automate each workflow, which is exactly what every person with a business wants to do.

At the top it also says you guys are doing agentic customer journey automation which spells that out more clearly. I'd love to hear two things: how did you get into the world of AI and then we'll go from there.

Sure. With my previous startup which folded around 2021-2022 due to COVID, I took a sabbatical to get my bearings. By the time I started thinking about what to do next, the AI era had begun with ChatGPT just about launching.

My co-founder Karma and I had been thinking about problems and a specific problem we had in mind coincided with the launch of OpenAI's ChatGPT. That changed our perspective and gave us an elegant way to build the solution we were thinking of.

I started my company in 2023, fairly recently. I've seen many so-called rappers in the space but we specifically built this AI agentic customer journey platform.

What was the specific gap you saw in the market where standard chatbots were failing and what was the light bulb moment where you realized the tech had matured enough to move from chat to where you're at?

Great question. It wasn't a light bulb moment but more of a journey. We wanted to figure out the problem we had in mind from my previous startup where we were selling health supplements online but were not converting enough despite a 71% completion rate on a 28-question test.

We found that eight out of ten customers would convert once engaged personally. Everyone had different reasons for not purchasing, and FAQs or call centers didn't solve the problem. We realized a high level of conviction and expertise was needed to solve it.

Fast forward a few years, the problem stayed the same. We wished we could replicate ourselves to engage customers and convert them. So Ziggman started to build sales agents with a persuasion angle, different from support bots where the customer leads the conversation.

We realized the solution had to fit into a much larger tech stack within marketing or growth organizations. It needed to be fully integrated with CRM, CDP, drip emails, and other tools. So it evolved from a conversational chat agent to a journey platform.

No worries about the long answer. That was good context. Things don't always come as light bulbs.

For listeners tired of buzzwords like AI agents and agentic layers, can you explain in plain English the difference between a standard conversational bot and the agents you are building?

The fundamental difference is autonomy. Conventional chatbots are scripted decision trees with no autonomy, asking fixed questions and following predefined paths, which leads to terrible user experience.

AI agents are trained with information and guardrails but have autonomy to make decisions not scripted or tested for. It's like hiring a sales person with general scripts and letting them handle different customers autonomously.

This autonomy allows AI agents to lead conversations naturally, unlike scripted chatbots that customers avoid because they can't decide how to get their questions answered.

Where do you think these agents fit in a team? Can they replace people in some contexts?

AI agents today do not handle entire tasks alone; there is always a human in the chain either before or after the agent's involvement. In deployments at Ziggman, humans engage before and after the AI agent's tasks.

In sales and marketing roles, many repetitive tasks exist such as responding to Instagram messages, sending nudges and reminders, and confirming meeting participation. These are tasks AI agents are good at.

Closure of business, especially for high-ticket items, still requires a human to negotiate and explain finer parts, but AI can shine in the earlier steps leading up to that.

There are many misconceptions about AI. How does your system handle sentiment-aware actions as mentioned on your website?

Previous chatbots stripped out signals like sentiment or mood detection. Current AI agents have human-level natural language conversations that allow detecting anger, frustration, urgency, and other sentiments.

Our platform constantly monitors conversations and updates signals like strong intent or urgency on the dashboard, enabling better responses and prioritization.

When did AI start picking up sentiment analysis better? We started with GPT-3.5 which had a high error rate, so we built strong guardrails. The big jump was with OpenAI GPT-4, which improved qualitative output significantly.

Reasoning models like GPT-4.1 take time and are not usually needed for conversational tasks where quick responses are important. Normal GPT-4 models suffice for most conversations.

When pitching to businesses, are they hesitant to hand over closing capabilities to AI?

We do not advocate handing over closing to AI. Most businesses want help identifying the best leads from many and focus their best people on those. AI helps distill leads and prep them for human follow-up.

Some clients initially expected AI to do full sales but later adopted a phased approach where AI handles parts of workflows, building confidence internally before expanding AI's role.

What was the hardest hurdle to get AI good enough to generate revenue and buzz?

We focused on delivering value even at the cost of scale and growth. The AI had to impact customer metrics, not just internal metrics. This focus built a culture of solving real customer problems.

Which channel is hardest to master: web chat, SMS, email, or social media?

They are practically the same in difficulty, each with nuances. For example, email has threads which are more complex than messaging threads, but these problems have been solved.

What keeps you up at night?

Every business has its own set of problems. We focus on finding commonalities to avoid reinventing the wheel and think about what kind of organization we want to be in two years and how efficient we will be at solving problems.

What is the company's main ethos and goals over the next few years?

Our ethos is very customer-centric and solutions-focused. We do not want to be a product company but a solutions company. Product is an outcome of solving problems, not the goal itself.

Our goal is to continue mastering conversations and focus on quality. There are many chatbot players but we want to be known for good conversational quality, like the difference between good and bad salespeople.

Your platform claims to boost conversions by 35%. Is that purely speed to lead or is AI better at qualifying than a human SDR?

We have a published case study with Meta and Nova Healthcare showing 35-40% better end conversions compared to parallel human teams running the same campaigns.

Other case studies include Bajage Auto in Spain with 23% better metrics after two months, Script Box with 25-30% better conversion, and an online stock broking firm where onboarding completion doubled from 10% to 24%.

Where do you think AI agents are heading in the next 12-24 months?

I think we've reached a plateau in inherent LLM ability. Future improvements will be gradual. However, smart developers are figuring out new ways to utilize AI beyond what has been done so far.

We are bullish on multi-cloud platform (MCP) trends and interoperability, but adoption will take time because businesses need proof of impact before adopting complex systems.

Capabilities may come quickly but effective adoption will struggle due to the complexity of business ecosystems and the need to show significant metric uplift to justify change.

In five years, do you think an entire entry-level SDR role could be replaced by AI?

SDR is the hardest role to replace because the starting and ending points of customer interaction are best served by humans. AI can automate many middle tasks but closure requires human elements.

I do not think there will be mass replacement of workers by AI. Humans are far smarter and we will coexist with AI as a tool with a larger impact than earlier tech evolutions.

Most businesses are not looking to replace humans but to increase throughput and grow the business rather than just cut costs.

Many people focus on how AI can help companies better serve customers and automate monotonous tasks without cutting staff, allowing current staff to focus on higher-value work.

My favorite personal AI tools include Text Cortex for writing agents, Crisp AI for removing background noise, and Gemini for non-coding tasks like image creation and document analysis.

You can find me at ziggman.ai, on LinkedIn as Digshant Dave, or email me at digshant@ziggman.ai. I share a lot on LinkedIn, so that's the best place to connect.

Please check out Ziggman AI at ziggman.ai and if you liked this episode, leave a like, comment, subscribe, and review on Apple Podcast and Spotify. Thank you all and see you next time.