Insurance Website Platform

Designed for
security, speed,
and fault-tolerance.

A solid architecture goes a long way.

Our recognition of the importance of website security, speed, and uptime
are reflected in our platform's fault-tolerant, high-availability architecture.

Automated horizontal scaling, graceful failover and recovery, and more are built in
with the aim of keeping your insurance agency website secure and online 24/7/365.

Keeping Everything Online


Our servers that handle website traffic reside in at least three geographically separate data centers, ensuring that if one or even two of those data centers lose power or internet connectivity, your website will remain online and available, and all of the incoming traffic will be routed to the healthy data center(s) until the unhealthy data center(s) recover and are ready to begin handling traffic again. This all happens automatically.

Additionally, our web servers are configured to scale automatically in the event of an unplanned traffic spike. This can happen if a website we host is linked to from a highly trafficked website without our prior knowledge, such as a news site (it's happened to sites we host before!).

In a sustained traffic surge event like this, if our baseline number of servers are unable to fully handle the load on their own, then more servers will be added to our server cluster automatically, and the traffic load will be spread equally among them. As soon as traffic subsides and the additional servers are no longer necessary, they are removed from the cluster to save on costs.

This benefits you not only because we can handle traffic spikes on your website, but because we can handle traffic spikes for other websites we host, which means another website's surge should not impact the speed or availability of your website.

Additionally, we've designed our systems to be flexible, allowing us to upgrade individual parts when the time is right, rather than having to upgrade the entire system all at once.

Keeping Everything Fast

In addition to designing and building a robust system with the ability to absorb unexpected traffic surges and gracefully handle partial service outages, we've also placed potentially slow and long-running processes on separate "worker" servers that are dedicated to these types of tasks. Processes like this should not be executed on the web application servers, because they can slow down website response times and otherwise impact server performance.

An example of a potentially slow task is email delivery. When someone submits a quote request form on your website, we store the form data in our database, place a worker task in a message queue, and respond to the website visitor with a "thank you" message right away, without any other processing. Less than a second later, one of our worker servers picks up the task from the message queue and processes it, resulting in the delivery of a notification email to you. Network trouble or other adverse conditions can cause email delivery to take several seconds, or even to fail and require a retry, and you don't want your website visitors waiting that long to receive a confirmation message that their request was received. Such a delay might reduce the visitor's confidence in using your website to communicate with you.

Keeping Everything Secure

We care deeply about security of all kinds (system security, physical data center security, data integrity, data privacy, etc.), and we work hard to keep our systems, software, and data safe from prying eyes and other malicious actors.

Data is stored in an encrypted format in our databases, in our object storage spaces, and on our file systems. LightRail, our proprietary software that runs our websites and administrative web applications, employs end-to-end encryption, meaning that all traffic between your computer and our servers is encrypted via TLS (also called SSL and HTTPS).

We have policies and processes in place for ourselves, and we promote these to our customers anytime the occasion arises, to help keep data secure as it is moved around. These policies include: never sending passwords via email, even to ourselves; using strong and unique passwords for individual online services (including our own); storing our passwords in an encrypted format; never signing in to our systems from untrusted computers; and much more.

You can read more about our security measures and approach on our security page.