We have some case studies based on Artificial Intelligence

Client Background

The Client is based in the USA. Client mainly works on providing HR services like consultation, contract staffing, training and development etc. to various MNCs in the USA.

Solution

Based on custom curated data and using ML pipeline we provided system where in customer upon providing 3 candidate profiles, we create a Job Description by extracting semantic info from profiles. Customers were able to reduce turnover time for Publishing Job Descriptions from 1 week to less than 5 min. Using provided system, Customers were able to find right skilled resource that specifically matches their requirement.

Problem Statement

The client’s lot of time was being spent on preparing job descriptions for job posts. To save time, the client wanted an automated AI solution wherein The System can generate Job Description Identify preferred skillset for Job Profile Finally, to publish the created job descriptions Client also wanted an UI where they can provide information about the Job requirement.

Tech stack used

Client Background

The client is based in San Francisco, USA. They are one of the leading software companies in the USA. The client has a chatbot product, It has ability to connect with variety of cloud applications like Google drive, box etc.

Solution

To check if the query made by a user is related to the files present in user’s drive, we train an NLP based AI model for each user. This model predicts to which file the query is related to. Later it predicts column name and row selector based on user’s query and give’s reply. For example, user makes query “what was sales on 12th September 2019.?”. Column name selected will be ‘Sales’ and ‘Date’ and row selected will be for 12 September 2019 date. To automate the process of saving attachments in the Gmail and save in Google Drive and Box Drive we created API which gets triggered on receiving an email.

Problem Statement

The client wanted enhance functionality of the bot. They wanted the bot to reply user’s queries based on data present in their cloud application account files (csv, excel, etc.) Functionality to automate the process of saving attachments in the emails from specific sender to be saved in a particular folder in Box and Drive.

Tech stack used

Client Background

The client is based in Bangalore, India, the client provides subscription-based recommendation service. The recommendations are provided for variety of sectors including Stocks, fantasy games (cricket, baseball, football, etc.) , commodity etc. The client has more than 100,000 subscribed members to their service.

Solution

As the client did not have dataset, we first had to extract historical data for the Korean baseball league. After scrapping the data, we preprocessed the data to remove unwanted values. Later, we used the prepared dataset to train models on variety of algorithms. We got 88% test accuracy using deep learning algorithm. As part of phase 2 we built a SaaS platform for the clients customers integrating the recommendation models. With our client was able to improve their profile by a factor of 4.

Problem Statement

The client wanted an AI based recommendation model to be built which would help in creating team of best players. This was supposed to be build for a particular Korean baseball league and client did not have any dataset available that can be used for training the AI model Also, they wanted frontend with UI where they can select teams for a match.

Tech stack used

Client Background

The client is based in New York, USA, and they are a leading real estate firm. The firm has inhouse team of 50+ real estate agents. They have been doing the business offline traditional way and wanted to upgrade making it an online business.

Solution

We delivered SaaS platform for real estate property listings having extensive features of promoting own listings using custom ad packages inbuilt for ensuring individual listings are getting maximum traction. Real estate property owners and agents were able to close deals at 60% more faster for listings posted on our site when compared with competitors, by using boost marketing strategy

Problem Statement

Client, a real estate agency, needed a next generation portal where agents and customers can interact and list real estate properties. Customer needed software which can allow agents, management realtors and building developers to post listings and allow seamless flow for property seekers to go over listed properties and reach out to their neighborhood specialist.

Tech stack used