Challenges

Rampant Product Customizations

On-prem EHS implementations had custom fields, tables, and workflows, resulting in inconsistent data structures and definitions across customers.

Troublesome Unified Data Ingestion

Ingesting data into a SaaS platform across a myriad of customized products was posing a big challenge.

Harmonization and Individualized Extensions

The ability to define common KPIs was tricky. So was customers’ desire to have their customizations included in the I&I layer.

Natural Language Querying

Being a new product, customers were demanding that the I&I offering support natural language querying.

Solution

A solution involving several individual building blocks was recommended and the PoC was developed.

Semantic Model Identification Engine

This extends a base semantic model by applying the individual nuances in customers’ implementations.

Extensible Data Model & Data Ingestion Engine

A metadata-driven ingestion framework capable of pulling data from customized product layers, supporting schema evolution, hybrid deployments, and scalable downstream analytics.

NLP-Based Query Engine

A query engine that dynamically adapts to each customer’s customized data model.

Intelligent Visualization Layer

An agent-driven analytics layer that automatically selects relevant KPIs, generates appropriate visualizations, and provides contextual insights and annotations instead of static reports.

Impact

Having successfully developed some of the key components of the solution, the company is now confident about driving this major transformation in its product portfolio.

The company is already engaged in active conversations with several of its top customers with live demonstrations of the envisioned I&I product offering. The initial responses have been highly encouraging.