CHATBOT
AI based Agent Assistant Bot with NLP capabilities

A leading university with multiple campuses and offering versatile courses needed a tool that can handle student queries 24/7, in multiple languages and without sounding like a bot. SSDB built AI based chatbot with natural language processing capabilities (NLP) that could interpret questions asked in natural language and could revert response in natural language.

  • Intelligent conversational workflows to handle common services
  • Discovering and delivering relevant, personalized and contextual information
  • Self Learning Engine with cognitive capabilities and robust feedback mechanism

For Administration - Attract the right kind of students with highest probability of success. Enhanced decision making to reveal priority programs. Early identification of at risk students, and prediction of pass rates. Improve student retention and check dropout rates.

For Students - Facilitate placement information. Make informed decisions on what courses to study and training based on evidence. Check for eligibility on scholarship, fee payment (instalment) options, and educational loans. Campus, Infrastructure and Faculty details.

Corrective Action - Timely notify to improve achievement levels and better grades.

MOBILITY

Having a native mobile for your business is no more a style statement, it’s a necessity. Your customers demand is to stay in touch and be able to communicate with you as per their own leisure and without having to deal with a human. You offer a huge array of products, which they wish to explore – anytime they feel like, and yes if they like it they order it as well.

Our case - a leading bottler of coca cola in South Asia was facing multiple challenges with phone support through which customers had to call and order products. Customers had to deal with the frustration of waiting a long time on call and would often go to the competitor. Due to this bottleneck, they also were not able to roll out offers and schemes to the customers. They wanted to expand and be able to reach out to all existing and new customers with complete product portfolio.

Our solution - we created a solution that was a combination of the native Android app, native iOS app and a smart backend web-based interface. Native apps were for the end user to browse products, see offers, place an order, make payment and track delivery. They could also add favourites and add reminders. The web-based backend was for the back office operations and fully integrated with ERP, it acted as a single window to monitor and manage entire operations related to customers, orders, payments, inventory, offers, production and delivery. We also created an android app for delivery team, this app was linked with maps so that exact customer location could be tracked and fastest route to deliver could be assessed. The delivery team could track how many bottles to deliver, how many empties they need to pick and what’s the payment due for the customer. Our solution was a huge success right from the launch as the users really waiting for it for a long time. Client’s business shot up by 500% within weeks from launching.

MACHINE LEARNING

What is it? Machine learning and artificial intelligence are synonymous with everyday life. Every business sooner or later is realizing its potential and rolling it for their customers. Machine learning not only makes a machine learn and adapt itself to the user actions & preferences but also is fast, accurate, and makes a machine capable of decision making on its own.

For Administration - Attract the right kind of students with highest probability of success. Enhanced decision making to reveal priority programs. Early identification of at risk students, and prediction of pass rates. Improve student retention and check dropout rates.

Our case - A globally leading player in legal contract management approached us with two challenges: how to eliminate the process of manual entry of contracts (with proper categorization and tagging of contents) and how to enable end users to access specific contents of the contract instantly.

Our solution - Our machine learning team got into action and built following

  • A machine learning based document extractor that can read and extract PDF contents, analyse the text, add necessary tags and segregate the contents to applicable headings. It also could read dates and set alerts automatically. All this done in split second.
  • A mobile app based chatbot, integrated with Watson, that became front face of the entire contract management system. The user could interact with the chatbot directly through voice or text and get any details for any of their contract. For instance, they could now ask chatbot – “When is the lease agreement with party XYZ due for renewal?” and chatbot would give the exact date in return.
  • Increased customer satisfaction and a huge reduction in customer on-boarding time were some of the obvious benefits for the client.
STATE OF GOODS REPAIR

Helping transit agencies maintain bus and rail systems in a State of Good Repair (SGR) is one of Federal Transit Administration’s (FTA) highest priorities. FTA recommends Transit Asset Management (TAM) practices to preserve and expand transit investments.

Our client was looking for better tools & infrastructure to migrate their existing facility inspection and SGR calculation procedures from a mix of manual + web methods to fully automated and secure platform.

Our case - Covering a service area of 5,325 square miles, our client is the US third largest provider of bus, rail and light rail transit, linking major points of three US states.

Our client, a public transportation corporation, operates a fleet of 2,027 buses, 711 trains and 45 light rail vehicles. On 236 bus routes and 12 rail lines statewide, they provide nearly 223 million passenger trips each year.

As a mandate, each facility needs to be inspected once in every three years to assess State of Good Repair by third-party independent inspectors. These inspections were carried out using pen and paper and then later notes were collaborated on the web to complete the inspection. This approach had some obvious challenges like -

  • Manual feeding of records lead to manual errors
  • No ability to collaborate photos with findings
  • Duplication of efforts, first capturing on paper then retyping on web
  • Loss/damage to paper in transit would lead to re-evaluation
IOT - PUT TO LIGHT INTEGRATION

Our solution enabled the client to conveniently and effectively add e-commerce fulfilment business to their portfolio and offer it as an extended service to its existing bulk clients.

E-commerce brings millions of orders per day, from something very small like a pen to smart TV – people buy everything online and the orders are delivered within a day if not in hours. This in turn puts a huge pressure on warehouses to pick, pack and ship the items ASAP.

Our Case - A leading warehouse in the US, having warehouses at multiple locations, handling bulk as well as e-commerce based business for its clients needed an end to end solution that could handle all warehouse operations – from unloading and receiving items to packing and shipping them out through shipping carriers globally. In the entire solution speed was essence and a major time taking activity was picking the right item ordered from millions of items stored in the warehouse. The question was how?

Our Solutions - After carefully analysing several options we figured nothing matches the speed of light; as a result, we implemented a light-based navigation system popularly known as a pick to light or put to light (PTL). In layman terms, each bin location (where the item is to be picked or has to be stored) has a light associated with it and each light has a unique ID. When an item id has scanned the light from where the item needs to be picked/stored turns on automatically and the store person can quickly locate it in no time.

Our solution enabled the client to conveniently and effectively add e-commerce fulfilment business to their portfolio and offer it as an extended service to its existing bulk clients. Now our client has a fully integrated application that runs on high-end handheld devices with touch screens and inbuilt scanner through which user does unloading, receiving, stowing, order fulfilment and ship out.

BUSINESS INTELLIGENCE

Business, no matter how small or big, churns good amount of data all the time. As the business grows so does the data and with that an implicit issue arises of managing, analyzing and visualizing what we have in our data and what does it tell about the future. Fortunately, with spread of advanced analytical tools, trawling through Terabytes of data is no more a challenge. SSDB team has experts in leading analytical tools like QlikView, Tableau, SSRS to name a few that helps organization understand their data faster and gets better decision making capabilities.

Our Case - A leading warehouse in the US, having warehouses at multiple locations, handling bulk as well as e-commerce based business for its clients needed an end to end solution that could handle all warehouse operations – from unloading and receiving items to packing and shipping them out through shipping carriers globally. In the entire solution speed was essence and a major time taking activity was picking the right item ordered from millions of items stored in the warehouse. The question was how?

Our Solutions - we got into all the tables, identified relevant tables, implemented an ETL tool that was capable of extracting the details needed from several distributed servers, and load them into a single server in a compressed state so that analysis was faster. Then using advanced BI capabilities of Tableau, our team built advanced dashboards that represented the call information based on demographics. Using our dashboard the client could visualize real-time details of the calls, could slice and dice the data based on geography and or provider and analyze Jitter, packet loss, latency, and overall MOS (mean operating score) of the calls.

Using our solution, the client was able to clearly identify areas of improvement and got a better handle on troubleshooting needed to improve call quality. Now, if customer complains about a poor call they could exactly look and tell what went wrong.