AI, Ontology and Relation Graph

How IQLECT enhanced Cisco's Customer Experience (CX) and Sales team efficiency by building an Ontology for their CRM.

BUSINESS CHALLENGE

The CRM team at Cisco had a tough time handling the number of issues raised by their customers regarding configuration and compatibility. They believed that the majority of such issues could be resolved by suggesting the right parameters to their customers as soon as they start typing the issue-related details.

Stream Layer

Batch Layer

About

SOLUTION

To address the second challenge,Cisco chose to use IQLECT‘s ontology creation app which provided a readable interpretation for router-based issues using ontology graphs based on domain knowledge. As ontology graphs are interconnected and interoperable, they play an integral role in addressing the challenges of accessing and querying data in large organizations like Cisco.

A few reasons why Cisco chose IQLECT:

• The platform enabled entity extraction from unstructured or semi-structured texts.
• IQLECT kept BangDB into action which is a multi-flavored, distributed, transactional, high performance NoSQL database suitable for heavy lifting.
• Presence of ML/IE layer to perform sentiment analysis which helps in suggesting a relevant solution to a user who is facing compatibility or configuration related issues by monitoring customer information.
• Use of dual machine learning models such as CRF and SSVM for collecting relevant information from the product and support documents and produce a structured representation in a semantically precise format.

Measures taken by IQLECT to solve Cisco's ontology problem:

• We created a repository of all the existing product/support documents of Cisco.
• We have kept the machines learning at work and train the knowledge base for product and support.
• We let the ML find and understand relevant parts.
• This enabled gathering of information, finding relationships between the entities and organising it in a structured manner.
• Produce a structured representation in semantically precise format - which is easily understandable and accessible.
• Allow other algorithms to make inferences.

CLIENT SPEAKS

“We were first introduced to IQLECT as part of the 2016 Cisco LaunchPad event where we collaborated on an additional use-case. We have been associated with them since early 2017.We realized their potential fairly early after which we on-boarded them under formal contract as vendor somewhere around Q2-2017.

IQLECT has worked on initial ontology project for Cisco product and support pages to help automate processes in the CRM team. Together we deployed the platform along with solution within Cisco in Q3 2017 and we are exploring deeper integration with the platform for more use cases. This has also been presented to Cisco leadership who are very impressed with their overall work.Our experience working with IQLECT has been very good, we value their product and offerings and based on our experience we’ve also offered their services to other service entities within Cisco who are looking towards solving similar business challenges.

Their approach towards solving data related problem is unique and that combined with their home-grown converged platform only allows users to enable complex use cases effectively and in short period of time. An integrated ML/IE layer gives our solution a much-needed edge.

The entire IQlect team has been very supportive so far and our experience working with them is very positive. We foresee lots of potential going forward and look forward to a much deeper engagement with them.