MongoDB World 2018: From Disruption to Transformation: Document Databases, Domain Driven Design, and Microservices at Travelers Insurance

1
From Disruption to Transformation
Document Databases, Domain Driven Design, and Microservices at Travelers Insurance
Michael Braasch & Jeff NeedhammongoDB World 2018

© 2018 The Travelers Indemnity Company. All rights reserved.
Introduction
Michael Braasch
2nd
Vice President
Travelers Business Insurance
@mbraasch
Jeff Needham
Senior Architect
Travelers Business Insurance
www.linkedin.com/in/jeff-needham
2

Innovation
3
Leverage new technology.
Redesign the way we manufacture, and sell our products and
services to improve our productivity and efficiency.
Anticipate how future customers, and our agent and broker
partners will access and interact with our products and
services.

Technology
Software is the differentiator…
Engineering (developer) skills and productivity are
arguably two of the most important commodities in
business.
IT is at the front lines of innovation and transformation.
Chance favors the prepared – you have to act.
4

Software Delivery Challenge
5
Significantly reduce the time from
business ask to production

Meeting The Challenge Of Time
Disruptions:
Business
Architecture
Development
Testing
DevOps / ProdOps
Where we are now?
Transformation
6

In The Beginning…
In 2012, we struggled
In 2013, Scaled Agile Framework was adopted
 Cross functional Agile teams were established
 Incremental delivery
 Lean processes
Built a simplified, single ‘Workstation’ for Account Executives and
Account Managers to view and manage their book of business.
Simplified and modernized technology landscape.
7

2012 Legacy State - Where We Started
8

2014 – ODS Of Federated Data
9

Workstation…2014
Benefits:
 Single View of Accounts and Policies.
 Restful Services.
 SAFe/ Agile Program:
- Cross Functional Agile teams.
- Product Council to prioritize and manage the backlog.
 Integrated data from complex Legacy systems.
Cons:
ETL dependent
Monolithic
- Central Web/App Server where all code compiled
- Single Database
Integration challenges - asked to transact in two places.
10

11
business ask to production.

Microservice At The Center
12

Disruption…2014 - 2017
13
Increase Flow
Reduce Lead Time
Faster Feedback Loops
Reduce Risk & Disruptions for Ops
Simplify & Optimize
Lean Product
Management
Achieving the effectiveness of Lean Product Management requires,
alignment from both “Business” & “Technology” value streams
BUSINESS
Domain Driven
Development
ARCHITECTURE
Modularity
(Micro Frontends,
Microservices, and
Event/API based apps)
Ownership &
Accountability
DEVELOPMENT TESTING
Automation
Self Service
CD Pipeline
Enablement with
Blue Green and
Canary Deployments
DEVOPS PRODOPS
Monitoring
&
Alerting

14

Prospect
Background:
 Manage the sales process – from Prospect to Customer
- Unify multiple legacy systems – phased approach
- Single view, common definitions, workflows and metrics
- Single source of truth
15

Prospect – Release 1
Database Objects:
16

Database Objects:
17

The Move To The Document Database
18

The Rationale For Change
19
Agility:
 Relational—Agile Mismatch - Get it Right the First Time vs MVP
 Simple One-to-Many Relationships - Multiple database objects
 Object—Relational Mismatch – Mapping layer required
 Volume, Velocity and Diversity – On the rise - exponentially

Change Impacts
20
Cultural:
 Relational has been the bedrock.
It’s still all about entities and attributes.
Agility is not instability.
 Information Architecture – Standards and Controls.
Data Models
Names
Validation

Change Impacts
21
Technical:
 Your Architect is your most important Developer-Advocate.
 Training – Not just for your Developers and Engineers.
 Embrace the shell - Move from .find to .aggregate.
 OpsManager - Brings Development DBA to the Development Team.
Accessibility to previously DBA-owned responsibilities.
Transformative – and lets the right people do what they do best.

Database Objects:
22

Before and After Code Sample:
23

Success With MongoDB
 First release using MongoDB, developers had no prior experience.
 2 major rewrites (and 2 previous releases) prior to moving to MongoDB.
 Changing requirements continued after migration, but MongoDB’s flexible
schema allowed for rapid development cycles.
24
 Achieved overall Workstation goal:
Significantly reduce the amount of time from business ask,
to production delivery.
8 Weeks
from start to finish
Significant Increase
in developer productivity
600+
lines of code eliminated

Velocity with MongoDB
RDBMS Growth

2013 – 40 tables in Production.


MongoDB Growth

Summer 2017 – 2 Collections in Production.

Present – 120 and counting.

75% of our services are writing, or have plans to write to MongoDB

Surge of new structures = we’re building more widgets in less time.
25

Transformation
With MongoDB, Microservices, DDD and Continuous Delivery

Single team coordination: one microservice, one database.

One pipeline for application and database changes.

Deployment time is now in minutes, not hours… or days.

Have more time to spend on integration challenges.

All new work in MongoDB.

JSON is the new SQL.
Still in pursuit of this goal!
Significantly reduce the amount of time from business ask, to production
delivery.
26

27
Thank you for your time.

28
Questions & Answers

29

MongoDB World 2018: From Disruption to Transformation: Document Databases, Domain Driven Design, and Microservices at Travelers Insurance

More Related Content

What's hot (20)

Similar to MongoDB World 2018: From Disruption to Transformation: Document Databases, Domain Driven Design, and Microservices at Travelers Insurance (20)

More from MongoDB (20)

Recently uploaded (20)

MongoDB World 2018: From Disruption to Transformation: Document Databases, Domain Driven Design, and Microservices at Travelers Insurance