Looking for our FDAta app? Click here
Data engineering and automation for medical device and pharmaceutical companies
A functional service provider specializing in
- statistical programming
- automated report generation
- process improvement with AI/ML workflows
With solutions designed to meet regulatory requirements
Designed to meet regulatory requirements.
From a leadership team with over two decades of experience in phamaceutical and medical device validation.
Still using that tool that was written ten years ago?
Maybe by someone who left the company five years ago? You're not alone. More importantly, we can help.
Contact usgemba - IPA: /ˈɡɛm.ba/ /ˈɡem.bə/
- - in Japanese business theory, the place where things happen in manufacturing. Used to say that people making products are in a good place to improve the process by which they are made.
datagemba - IPA: /ˈdeɪ.t̬ə ɡem.bə/
- - in American business theory, the company to turn to for automated data workflows
Tools we use
-
Async Python Stack with:
- Task queues with
CeleryandRabbitMQ NumPy& relativesSciPyandstatsmodels- Machine learning with
sklearn - Computer vision AI with
pyTorch - Among many others
- Task queues with
-
R Stack with:
- Base
Randtidyversedepending on project needs (if any) - Visualization with
ggplot2 - Automated report generation with
rmarkdown - Linear mixed models with
lme4, Bayesian modeling withrStanorbrms, among others shiny
- Base
- SQL with
postgresql - Frontend visualization with
d3.jsand vanilla JS - Version control with
git Jinja2+Quartfor small projects orDjangofor larger onesWe're occasionally asked about SAS, to which our response, paraphrasing Jay-Z, is:
If you havin' stats problems I feel bad for son, I got 99 problems but SAS ain't one
- We do help organizations transistion away from SAS, but do not find it to be a good use of our time to maintain or create anything new in SAS.
- We often compare SAS to COBOL, and if you're wondering what COBOL is, you're already starting to see the point.