The Munich startup E-Bot7 develops and integrates artificial intelligence into companies' existing customer service. The system analyzes incoming messages, routes them to the appropriate department, and provides support agents with smart response suggestions during their daily business. This reduces processing time by up to 80% and automates recurring inquiries. We spoke with Xaver Lehmann, one of the founders.
E-Bot7 combines two trending topics—bots and artificial intelligence (AI). How do you see the market potential for this area?
The market potential for artificial intelligence solutions in customer service is enormous. Today, most CRM systems are completely outdated. Customer service representatives spend most of their time answering repetitive questions instead of focusing on more sophisticated issues. This inefficiency can be remedied with the help of AI and chatbots.
The market potential for virtual chatbot assistants is also enormous. Forrester predicts a 100-fold increase in market volume by 2020. Our initial target markets are the telecommunications industry, insurance companies, and banks. The market volume corresponds to approximately €25 billion.
Can you give a brief overview of the different bot solutions?
There are various technical approaches to building a bot. Most bots found on Facebook Messenger or as chat windows on websites are so-called "rule-based bots." These are usually built from existing frameworks and manually fed with data—questions and answers. This data forms a so-called static database, since the questions and answers used do not change. These bots work well as long as the customer asks a question that already exists in the database.
However, if a question is asked that doesn't exist in the database, the customer receives an error message. At the end of the day, the service agents have to go back into the system to read the error messages and manually insert the appropriate answers. Firstly, it's impossible to cover all possible customer questions with a static database, and secondly, no company wants to spend additional resources training the bot at the end of the day.
So what can you do better than other bots?
At E-Bot7, we've solved precisely these problems by using artificial intelligence and neural networks to understand the meaning of incoming questions. To ensure our artificial intelligence doesn't pass on incorrect answers to the customer, we offer a hybrid agent and AI solution.
Hybrid agent and AI solution
What exactly does that mean?
The company initially selects a required level of certainty, e.g., 98%, which allows the company to determine how certain the AI must be before automatically sending a response. As soon as a customer asks a question, the AI provides a suggested answer.
If the AI is confident that the suggested answer fits the question, the AI automatically forwards it to the customer. However, if the AI is only confident about 95%, for example, the AI forwards the suggested answer to an agent who reviews it.
The agent can now trigger three actions. First, the agent can submit the suggested answer and continue training the intelligence. This means that if the question is asked again, the answer has a higher probability. Second, the agent can edit and submit the suggestion itself, thereby further training the system by adding the edited answer to the knowledge base. Third, if a completely new question is asked, the agent can compose, save, and submit a new answer. This process builds a database that continuously adapts and learns as the agent operates. This allows the AI to accumulate new knowledge with each interaction. At the same time, the AI is controlled by the agent through "supervised learning."
“The results surprised us”
Your AI seems to learn quite quickly. Can you provide some numbers on this?
When it comes to AI and machine learning, it's important to remember that the system needs data and time to learn quickly and effectively. Accordingly, AI learns faster with 10,000 queries per day than with 100. The results of our technology surprised even us.
In one of our case studies with a large telecommunications company, our bot was already sending more requests (61%) than the agent after two months. Our focus on German helps us train other European languages relatively quickly—within four weeks.
What are your three ingredients for success?
- We have a very ambitious vision. We want to be the European market leader in AI and machine learning applications within the next three years. This goal motivates us every day to work even harder and with more focus.
- A strong team you can trust and share success with. You can successfully start a business and grow together.
- Perseverance and courage—we don't let critics and naysayers discourage us. There are plenty of them in the B2B business. We take criticism to heart and learn quickly.
What is a tricky topic for you at the moment?
Our team is growing very rapidly. Therefore, it's currently a major challenge to find suitable employees who fit perfectly into the team and have the necessary drive, like us founders.
Focus on added value
In your opinion, what is the decisive factor for sustainable corporate development?
In our opinion, there is more than one decisive factor for sustainable corporate development. However, one of our most important success factors is our focus on our value proposition. We must be clear every day about who, how, when, and where our software adds value. Without this awareness and conviction in our own product, we cannot achieve long-term success.
Fail forward – what was the most instructive mistake you made?
During our first investment round, we were overconfident that everything would go according to plan. At the last minute, one of our investors dropped out for no reason. Fortunately, we had a very strong network of other investors at the time who supported us. We learned from this that a deal isn't complete until everything is signed. (Editor's note: Here you can find out more about the financing round).
What advice do you give to other founders?
It's worth working 24/7 for your dream.
“Munich is the ideal city for startups”
And what advantages does Munich offer a founder in your opinion?
Munich is the ideal city for startups. Apart from the high rents, Munich has the best infrastructure for startups in Germany. Firstly, it has very strong universities such as the LMU and TUM with their entrepreneurship initiatives, such as the MakerSpace or the LMU ECHere, founders with complementary skills can find each other and begin implementing their idea.
In addition, there are many VCs and accelerators in Munich, which could provide the first venture capital. Finally, Munich is simply the most beautiful city in Germany.
Of course, we agree with Xaver and thank him for the exciting interview and the insights!
