Hyperspect AI

Hyperspect is a revolutionary tool for sales, growth marketing, and lead generation teams. It automates prospect research and generates personalized opening lines for sales outreach, seamlessly integrating with existing sales tools.

Featured Image

# The Story

Our client came up with an idea which would supplement the efforts of the sales team and revolutionize the growth marketing promoting lead generation. Here came the need of a technology partner who would carve a customized solution to this. We, Octify Technologies, rolled up our sleeves, researched and developed the product in and out, assisting our client to realize the product they aspired for, viz. Hyperspect. We developed a robust AI Engine which aims at helping SMBs accelerate sales by leveraging the power of machine learning and reduce the manual efforts to a minimum. At the heart of the service, would lie the tendency to produce high volumes of relevant data which would help in sales outreach and scaling.

# The Problem

The initial proposed goal was to architect a product never-made, from scratch, acquire the users, test the market response and attract investors which would help in converting the idea into a full-fledged B2B product. We had a strict timeline to build a reliable working model, which would mimic the actual product. And we published it within 6 months, saving the two full months from the projected timeline of 8-10 months.

# Our Solution

We embraced Agile methodologies using SCRUM project management methodology and laid out the project planning. We only picked up the latest and efficient technologies to wire the finished product. The entire application’s architecture we curated, included backend based out of AWS Cloud and that too with higher scalability. We used NextJS to develop user interaction and AWS Lambdas for the API System. Since it is an AI application, the database is the vital thing in this project for which we relied upon Mongodb. At the core of the project, we created an algorithm, which we named Master Personalizer. The algorithm uses GPT3 OpenAI Models to process the data. Alongside, we developed some business applications for the client like workflow automation tool, a highly interactive dashboard to take control over the data and to analyze it. Besides multiple amendments and incorporated, the complete aforementioned tech stack was developed way before the stipulated time.

# Mutual Benefits

First, the client benefited from the ahead of time delivery as they had enough time to make a suitable proposal. Second, the idea-in-action attracted the interest of business evangelists and investors they reached. The robustness and scalability of the product proved to be the game changer in the manual world of B2B applications. The anticipated benefits of the application and that too AI driven helped in raising the funds way more than expected.

# Fruitful Results

The application thus developed not only benefited the client but also helped us in adding another feather to our hat of diversity of solution development. Though there are many service providers out there, undoubtedly having big names and market capture, we got the full responsibility of developing the rest of the features and scale the app for the massive user base. Hence, we can proudly say that being a startup with limited yet enough resources, we can punch way above their weight class - be it the tech know-how, delivery or performance.

# Architecture & Technologies Used

The Hyperspect.ai application is developed on top of a wide variety of cutting-edge technologies to power its comprehensive feature set.
In order for us to construct the application quickly and easily, we decided to use Javascript technologies. The Material UI and NextJS Framework are used to build the front end. We make use of Google Firebase for authentication service. We are utilizing AWS Lambdas in the backend for serverless computing since we were quite clear from the beginning that we needed serverless solutions. To make code simple to manage and deploy on lambda, backend APIs are written in NodeJS and delivered with Serverless Framework. The backend design is built on the idea of microservices, and RESTful calls are used for backend and frontend communication. We decided to utilize MongoDB since this application depends on data and the database is constantly being read from and written to. The schema of MongoDB is quite versatile. Iteration cycles are made easier by adaptable schemas that enable fast adjustments as and when needs change. A partial failure while writing several documents to the database is one potential outcome. MongoDB gets rid of it via transactions. MongoDB serves as a one-stop shop for all of our issues. A quick and reliable queuing system called SQS is used to create jobs that are then consumed by the algorithm MasterPersonalizer. When an action is executed by a user at the front end, a job that follows the SQS producer-consumer pattern is deployed in the background. Automation has been used for every step of deployment, scalability, code error checking, code testing, and error collection. Our programs are installed on AWS, and we make extensive use of their services. Docker containers are used to deploy our apps on AWS ECS auto scaling clusters. We use CircleCI for CI/CD, which causes deployment from release branches. Application design is both secure and robust enough to manage several concurrent connections.

# Technologies Used

React JS, Redis, Nodejs, Ruby on Rails, Postgres, Ansible, AWS