Back
Crisp

Crisp

#crisp

One customer support platform to delight your teams and customers. Used by 1/2M people worldwide.
Launched June 15, 2015
Load previous page…
lower #crisp api system CPU load on 14 servers by 15% by adding a LRU cache on the common authorization check done on all routes, hitting MySQL everytime and now cache hitting most of the time
improve performance of #crisp api by caching user authorization grants for 20s in a LRU cache, avoiding roundtrips to our MySQL server for every single API request
improve #crisp chatbox quick replies that were deployed yesterday, allowing to restore custom quick replies anytime
Late-night work on #crisp
migrate all #crisp VMs running Node microservices from shared VCPU to dedicated CPUs so that micro-bursts from neighbors do not kill us at peak time
implement configuration in #crisp for upcoming predefined messages for the chatbox
start working on finishing up #crisp chatbox predefined messages feature
fetch more logs for recurring #crisp api downtime and find more potential reasons but still unsure
implement a request logger to #crisp rest api http proxy in Rust so that I can catch the request causing a runaway event loop issue in #crisp Node.JS rest api
release new version of #crisp REST API caching proxy, after almost 9 years to modernize code and help fix recurring #crisp API downtime — github.com/valeriansaliou/blo…
do improvement on oss project for #crisp
migrate Rust open source project dated back from 2017 to modern dependencies so that we can hopefully attempt to fix #crisp recurring API downtime that's still not fixed
design upcoming #crisp conversation Flash mode (fully agentic conversation)
eat ton of unhealthy chocolates for some comfort after 2 full focus days attempting to resolve cause of repeated #crisp api downtime
request delisting of #crisp emailing domain on.crisp.email from Mozilla Public Suffix List as its making yearly SSL certificate renewal VERY tedious as it requires to open a ticket with Sectigo support and a human check, which takes a ton of effort and time
Update campaigns statistics API #crisp
resume work on fixing #crisp potential reasons for repeated API downtime
sign off after a non-stop 13 hour work day on #crisp managing outages
address some CVE in #crisp backend dependencies
observe another new backend issue on #crisp new translation service, going out of memory so we need to partially rollback feature
Home
Search
Messages
Notifications
More

Are you sure?