Smart Hiring Online Advisor
Conversational AI is a massive step in the new era of digital transformation. The main task of it is to build a missing link in the chain of communication between people and technologies. Conversational AI enables technologies to understand natural language requests and answer them as humans would. Trust in Conversational AI helps businesses increase customer engagement and satisfaction through personalized experience, reduce operation costs and speed up the entire business process.
Our client is a European tech startup. The startup was built around the idea of creating an online advisor platform for people who are starting their own businesses and hiring people for their teams. Such young entrepreneurs might have a great business idea but lack suitable personnel to perform. For example, an entrepreneur has decided that he\she wants to be well- recognizable and decided to create a brand book. At what type of professional should the entrepreneur be looking at this point? Web designer or graphic designer? And can they avoid hiring a marketing director or brand manager at this stage? Our client’s goal was to help people get the right answers to such questions and find the exact professional they need in their team, not spending extra money on salaries for professionals they don't need right now.
Solution
Online advisor platform powered by Conversational AI chatbot and recommendation engine
How it works
Our mission was to provide client with a cutting-edge AI solution that enables customers to get human-like answers to their questions.
The user describes the need with a free text message
The bot relies on natural language understanding (NLU) engine to understand the message.
With the help of natural language generation (NLG),
the bot conducts the communication and asks additional questions.
After the communication, the bot gives the User recommendations of who can help to complete the task.
Our challenges:
The client did not have an AI expertise
The customer had only a business background, so our team had to manage the technical side of the project.
Non-stable workload
As our client was a startupper, the project had neither a fixed schedule nor a budget. In such circumstances, we could not have a stable development process and worked with long breaks between project stages.
There was no dataset for Machine Learning
Before starting the actual development process, our team managed to create a reliable dataset from scratch in tight cooperation with the client. We invented possible conversation scenarios to make the bot well-customized.
Project stages
In communication with the customer, we realized that the ideas are feasible to come try. We explained to customers how the bot would work, designed solution architecture, and developed a roadmap.
We decided to test the bot first on limited subjects. So, we validated the concept and AI capabilities on the requests related to the marketing
domain.
Our AI engineers upgraded the bot to support multiple domains and topics. Together with the client, we created a knowledge base of possible users’ requests and dialog flows.
We launched the bot using a progressive roll-out approach. We gradually directed traffic to the bot to ensure it works well with real users.
Now we are at the scaling stage. After analyzing the bot’s performance with the real audience, we further upgraded it and made the bot even easier to interact with. Satisfied with the result, the client is now considering expanding the functionality to add profiles of real people to contact with, not only showing users the job titles.
Summary
