Forethought: AI-powered customer support software development

Forethought’s generative AI uses Large Language Models (LLMs) to maximize efficiency, customer satisfaction, and ROI for customer support teams. Now your support representatives will spend less time-solving clients’ problems as AI-powered tools suggest all possible solutions to each issue. This is an AI customer support software for providing high-quality customer care to multiple channels using one platform with a wide range of integrations. The company offers four products for a more qualitative and automated customer experience: Solve, Triage, Assist, and Discover. We collaborated with Forethought while working on Assist and Discover. We enjoyed sharing our expertise and work with clients whose shareholders are world-renowned stars.

Services:

Industry:

MarTech

Location:

United States

Business Context

Cooperation between Geniusee and Forethought involved two of their projects for customer service automation: Assist and Discover.

“Assist” aims to make agents work faster, easier, and more efficiently, suggesting responses generated by AI and machine learning as LLMs (large language models) during agent-customer conversations.

“Discover” is tailored as a tool that optimizes workflows, tracks ROI and performance, and recognizes and nails gaps to solve customer issues. Of course, it’s also AI-powered customer service.

Key challenges:

Integrate multiple services for the Chrome extension and AI-powered customer support tools to assure alignment via various channels for a better customer service experience and faster problem-solving;
Create an analog of cross-platform services not just for organizations so that we can offer our services to a wide range of clients who use different software, frameworks, and programs.

Work approach

Constant self-learning

AI customer service required working with a diverse range of integrations with different systems and frameworks to ensure the correct AI-powered tools’ connections. To satisfy our client's needs and, eventually, improve the end customer experience, we had to adjust to a rapidly reshaped landscape.
Outstaffing

We collaborated on this project through the outstaffing model, meaning our team worked alongside the client's team. Well-designed cooperation between Geniusee and our client, coupled with deep trust, based on the wide knowledge base of Geniusee as a service agent, enabled this type of cooperation.
Start small to gain more

MVP, POC, prototypes, and demo versions—it’s always more efficient to test your product before implementing the desirable features, especially with AI technology. The price of a mistake is you spend years and years on the development of a product that users don’t like.
SCRUM approach

SCRUM allowed us to manage the software development process wisely and fast. We set the scope, schedule, and budget accurately from the start. This permitted us to ensure accurate resource prioritization and efficient usage.

Process

Forethought was looking for someone with significant expertise in integration services and AWS services. As an AWS Advanced Partner, Geniusee was a top-notch choice. We started our collaboration by working on the Assist project and then moved to Discover due to changes in the client’s priorities.

We managed our development process according to quarter goals and moved onto the software development lifecycle using SCRUM. A wisely chosen approach to project management allowed us to stage each sprint with an appropriate set of developed requirements and features. This ensured even distribution during all development phases.
We chose React because of its open-source framework with a wide range of free tools and libraries. Plus, React is one of the most popular frameworks for frontend development, as it assures smooth and easy system maintenance, making it quick and responsive. Our team also assisted in working with AWS services, e.g., Cognito for authorization.
Our goal was to ensure smooth connections of the AI-powered tools with Forethought’s end users: organizations and their support teams. All of them use different software, frameworks, and tools.
We didn’t have access to the initial system code due to security regulator requirements, which made our task more complicated, yet more exciting.
We closely watch the quality of deliverables and set the quality requirements from the very beginning with the client and test every part of the tailored services. Developers are responsible for code quality, so code reviews are vital. We also tend to use unit tests. In this project, Geniusee QA specialists completed manual end-to-end testing.
Post-release support and maintenance are not vital but necessary if you want to succeed. We continue cooperating with Forethought to ensure smooth workflow, fix bugs, and remove legacy code. No less important are system updates and the implementation of new features according to users’ needs and market trends.

Project tech stack

Various integrations with different services, platforms, frameworks, and software require great expertise in API connections. By the way, this was one of the reasons why Forethought chose us. We also worked with other AWS services and used React as an open-source framework with various free tools.

Cognito

Elastic Kubernetes Service

Geniusee team

Forethought approached us with a need to empower their team with specialists in integrations, yet our collaboration spread much further. We worked together based on the outstaffing cooperation model.

Product ream

Project manager

Development team

Frontend engineers

Oleksandr

Frontend developer

I would love to recognize their high professional level of company management. We had no overtime and precise estimates. There was no pushing so working with this client was comfortable. Company management succeeded in developing an active listening environment with an atmosphere of openness. It’s clear from the moment of onboarding that your thoughts and comments, feedback, and feelings are essential.
Daria

Project manager

I enjoyed working on such a fascinating project with the Forethought team. Forethought was very responsive and reacted positively to our suggestions and recommendations, as we had a sincere and transparent relationship. During the development process, we were able to gain a deeper understanding of the requirements and needs of the clients based on a deep dive into the context.

Are you looking for full cycle product development?

You found what you were looking for. Just fill out the contact us form.

Features

Chrome extension
The Geniusee team delivered a Chrome extension for an AI-powered tool to enable support representatives to use one platform connected to many different services they use daily.
We provided a soft bar with a feed, notes, favorites, search, and automated suggestions for ticket answers to clients based on artificial intelligence that goes deeply into the context of customer questions.
Tasks templates
Admins may add tasks to Jira and assign them to service agents (certain customer support representatives) using templates. This allows us to work faster and avoid manual mistakes with self-service.
Process automation
We strived to make support processes faster and easier, so we tailored data extraction according to the problem described in a ticket. It also depends on the chosen board. Integrations with certain products empower process automation. This is based on the main AI-powered functions that the Forethought team delivered, of course.
Smart notes
We developed a tool that allows support representatives to create work notes using a wide range of Forethought possibilities. End users (service agents) may create public and private notes and templates. Suggestions empowered by multiple variables are also presented.
Admin analytics
The Geniusee team tailored the analytical system in the admin panel. We delivered an interactive dashboard with the most necessary features and service reports. Admins (PMs, delivery managers, or any other responsible person in an organization) may check ticket quantity by agents and clients, amount of jobs done, and many more. This allows the use of all possibilities of AI for customer satisfaction and company profitability.
Autocomplete
Together with the Forethought team, we delivered a possibility for support representatives to make their work process faster and easier. Users may write an answer to a client, and the system makes suggestions and hints based on text context. This service is based on machine-learning services.

Results

A Chrome extension for customer support used in various industries from e-commerce to FinTech.
An AI-powered tool that connects with the most popular software systems and frameworks.
Integrations with tools, services, and platforms that support teams use daily.