Fabricks Alkemy™
We've built a solution to the practical challenges and pitfalls of machine learning (ML) projects - Alkemy is a Python framework delivering accessibility and flexibility at every stage of the machine learning lifecycle.
ALKEMY
How it works
Fabricks Alkemy removes critical blockers and keeps your ML project oriented towards delivering value. It paves the way for teams to find more open lanes. Here's how we do it:
SET UP
PROJECTS
Establish your environment for model development.
BUILD
DATASETS
Create any number of training datasets to test new model features.​
RUN
EXPERIMENTS
Build a variety of models to test performance, including benchmark models to beat. ​
MANAGE
VERSIONS
Keep the history and progression of your code, configurations, and model artifacts.​
DEPLOY AND
OPERATIONALIZE
Create a deployment for your best model so that it's ready to score live data.
ALKEMY
Why Fabricks Alkemy?
Data scientists have an ecosystem of data analysis and ML libraries at their fingertips, making it easy to pull together a dataset and establish a baseline; but very quickly they find themselves spending time on everything but data science.
​
Having built-in ground rules for keeping datasets and model artifacts structured and managed consistently helps keep data scientists organized and working as a team focused on the big prize. That's why we built Fabricks Alkemy.
Benefits:
STRATEGY
Fabricks Alkemy provides a framework for aligning model development with critical company objectives.​
​
ACCESSIBILITY
Fabricks Alkemy steers users towards simplicity with a structured approach for each stage of the ML lifecycle.
FLEXIBILITY
Seasoned ML pros are free to think outside the box and manage the tradeoff's of complexity as they see fit.​
​
​
VISIBILITY
Fabricks Alkemy captures the data you need to benchmark modeling experiments, monitor production performance, and keep stakeholders informed.​
INTEGRATION
Sync model outputs and project state to your data warehouse and orchestrate workflows with Fabricks Alkemy plugins for Apache Airflow, a crowd favorite.
​