Overview

EverGood Insurance is a US-based insurer running policy sales and customer service through its own digital channels. Support executives handled every query in its customer chat, and most of what reached them were routine questions about policy coverage, which meant searching long policy documents while the customer waited. Bacancy Technology built a RAG-based assistant on Databricks that answered those questions in the chat itself, drawing on EverGood’s own policy documents and citing the section behind each response. Once it was handling most of the routine queue, EverGood moved support executives to other roles in the business.

Technologies Used

Databricks
FastAPI
Docker
Python
React
Git

Project Highlights

checkmark

AI assistant added to the client's existing customer chat platform

checkmark

Majority of policy queries answered without a support executive

checkmark

Answer generation based on the client's own policy documents

checkmark

Support team involved only on queries outside the assistant's scope

The Challenges

1

Every customer query went to a support executive, including routine policy questions that repeated across thousands of chats. As the customer base grew, so did the volume the team had to work through.

2

Executives had to search the whole document for each query while the customer waited, and the same question could even receive different answers depending on which executive handled the chat.

3

The client was automating repetitive work across the business. A large share of the support team's time went to routine policy questions, which was the work the client wanted automated first.

4

The support team was available only during business hours. So, the queries that came outside those hours had to wait until the next working day which resulted in delayed response.

Solutions by Bacancy

1

Our Databricks developers built a retrieval assistant into the client’s existing customer chat. Policy documents were ingested into Delta tables and indexed with Mosaic AI Vector Search, and the assistant answered routine policy questions directly, without routing them to a support executive.

2

Bacancy Technology’s data engineers parsed the policy documents clause by clause. Retrieval returned the clause that governed the question, and every answer carried that clause as its source, so the same question returned the same answer each time.

3

For the initial testing, our AI/ML developers scoped the assistant to the routine policy questions that made up most of the queue. Anything outside that scope was directed to the support team, who then took it over the call.

4

Our Insurance IT services team deployed the RAG based support assistant on a Databricks model serving endpoint and connected it to the client’s React chat through a FastAPI service. It ran independently of the support team’s shift, so queries received outside business hours were also getting answered the same day.

Core Features

checkmark

Policy document ingestion into Delta tables under Unity Catalog

checkmark

Clause-level retrieval with the source attached to every answer

checkmark

Escalation to a support executive on any question outside scope

checkmark

Integration with client's existing support platform with model serving

No. of Resource

04

No. of Resource

Time Frame

Sep 2025- Mar 2026

Time Frame

Project Snapshot

EverGood Insurance Dashboard

Outcomes

70% of chat volume was being handled without a support executive, across the query types in scope

The assistant answered from the current policy wording, so revisions applied to every response once published.

80% of the support executives were moved to other roles within the company without impacting the support workflow.

Queries placed outside business hours were also getting managed on time, resulting in more active customer traffic.

Each answer cited the section of the policy document it was drawn from, giving the team record of what the assistant told customers.

The assistant's scope widened across query types after launch, with the support team handling the remainder

Technical Stack

Frontend React
Backend / API Layer FastAPI
Storage & format Delta Lake
Governance Unity Catalog
Vector index Mosaic AI Vector Search (Standard)
Generation Databricks Foundation Model APIs
Serving Databricks Model Serving
Evaluation Mosaic AI Agent Evaluation
Orchestration Lakeflow Jobs
Containerization Docker
Scripting & automation Python
Version control & CI/CD GitHub Actions

Experience With Bacancy

Optimize Your E-Commerce Workflow for Maximum Efficiency

iconWatch the video

Fintech Platform Scalability & Growth Optimization

iconWatch the video

Legacy Code Upgrades for Modern User Experience and Enhanced Data Analytics

iconWatch the video

Empowering Retail Marketing With Seamless User Experiences

iconWatch the video

Migrating To Laravel From Core PHP With Twilio Integration

iconWatch the video

Upgrading PHP Laravel version while scaling the Admin Dashboard

iconWatch the video

Optimizing CRM Performance With Data Management and Salesforce Integration

iconWatch the video

AI Solution Engine for Improved Performance and Error Handling

iconWatch the video

Enhancing Managed IT Services With Streamline Automation Process

iconWatch the video

SEO and Bulk Data Upload Solutions For An E-Commerce Healthcare

iconWatch the video

Elevate Your E-Commerce with Scalable UI and Enhanced Calculators

iconWatch the video

Simplifying Annuity Rate Management for Complex Calculations

iconWatch the video

Addressing Null Safety, Code Structure Issues, and Manual Deployment Issues

iconWatch the video

Optimize Task Management With Advanced Tracking

iconWatch the video

Revolutionize Applicant Tracking: Advanced Image Optimization Techniques

iconWatch the video

Cost-Effective Media Management Solution and Resolving Data Upload Issues

iconWatch the video

Enhance Your Management Service with User-Centric Innovations

iconWatch the video

2500+ Projects Experienced Innovation with Bacancy!

Get access to an experienced team of developers and engineers from Bacancy, handpicked to ace your goals. Kickstart within 48 hours, no-risk trial.

Book a 30 min call

14+

Years of Business Experience

1458+

Happy Customers

12+

Countries with Happy Customers

1050+

Agile Enabled Employees