Pipelinepharma website development on python Django

Stack change form react + nextjs to python Django success story

 

Pre - history

Being contacted by creators of largest B2B pharmaceuticals marketplace presented with the unique challenge, which was more than happy to take. 

 

Initial challenges

As many of fast growing  businesses these days, client had lots of challenges. Mos notable ones were:

  1. Slow releases - Time span from idea to delivery was around 2 weeks. More for larger scope ideas
  2. Data security - GDPR compliance and sensitive data security was a concern of a founders, but not that much of a software developers
  3. Platform stability - downtimes and crashes were part og daily life, with observe availability of 82%
  4. Platform speed - average operation took around 1.2 seconds to perform. SEO vise platform scored around 30% in terms of performance, and  70% in terms of SEO.

 

Switch to python Django

After failed attempts to stabilize and improve situation on existing web stack, decision was made to change stact to python Django. 

This took nearly 4 months to release, and first working MVP. Results were quite good - page speed insights reported all metrics 90+.

 

Stability and transparency

Together with new stack we implemented alerting and monitoring based on loki, prometheus and grafana. This allowed us to know exactly what went wring where, and quickly fix it, allso, measure operation times and assure that 99% of client requests would be processed under 250 ms.

 

0 downtime automated deployments

With new release we allso conteinerized entire stack, allowing to deploy and scale entire solution. Deployments got automated, so that on merging code changes, workflow would deploy new version of the code with 0 downtime of actual platform in under 3 minutes. Test would be ran, and security checks ensuring that staging testing and tests ran sucsessfuly are in place, ensuring high stability of production environments. 

On each release to production, main stake holders receives slack messages on what was changed, allso corespojnding Jira tickets gets updatd, allowing further actions to moneterize newly added features. 

 

12 Releases a day

Having proper stack and good workflows allowed client to have 12 releases a day done on averagre, unlocking new experimenting potential matched with great speed of solutions production.

 

Design system

Thanks to chosen tech and solution architecture, all visual elements became components, and got reused in entire platform, allowing rapid releases, and uniform look and feel all over the workflow

 

Backend reporting

In order to improve user funnels, we implemented cookie-less back end reporting system, allowing 100% precision, full complience to GDPR and still ability to understand exactly how users interact with a system, and which advert chanels works best.

 

Conclusion

Despite being intense and hard experience, Pipelinepharma was allso most revarding one. +30% of organic trafic increase, new 100% seo scoring platform with A+ security headers stands as a proof of that.