Databricks for MLOps Presentation (AI/ML)

MLOps with
Databricks
Presentor :
Tanishka Garg

Lack of etiquette and manners is a huge turn off.
KnolX Etiquettes
 Punctuality
Join the session 5 minutes prior to the session start time. We start on
time and conclude on time!
 Feedback
Make sure to submit a constructive feedback for all sessions as it is very
helpful for the presenter.
 Silent Mode
Keep your mobile devices in silent mode, feel free to move out of session
in case you need to attend an urgent call.
 Avoid Disturbance
Avoid unwanted chit chat during the session.

1. Understanding Databricks
2. What is MLOps?
o Components of MLOps
o Benefits of MLOps
3. MLOps with Databricks
o Why Implement MLOps with Databricks?
o Development stage
o Staging stage
o Production stage
4. Components of MLOps with Databricks
5. Demo

Introduction to Databricks
 Databricks is a unified, open analytics platform for building, deploying, sharing, and maintaining enterprise-grade data, analytics, and AI
solutions at scale. Most of the tasks include:
o Data processing scheduling and management, in particular ETL
o Generating dashboards and visualizations
o Managing security, governance, high availability, and disaster recovery
o Data discovery, annotation, and exploration
o Machine learning (ML) modeling, tracking, and model serving
o Generative AI solutions

What is MLOps?
 MLOps stands for Machine Learning Operations.
 MLOps is a core function of Machine Learning
engineering, focused on streamlining the process of
taking machine learning models to production, and
then maintaining and monitoring them.
 MLOps is a collaborative function, often comprising
data scientists, devops engineers, and IT.

Benefits of MLOps?
 Efficiency:
o MLOps allows data teams to achieve faster model development, deliver higher quality ML models, and faster deployment and
production.
 Scalability:
o MLOps also enables vast scalability and management where thousands of models can be overseen, controlled, managed, and
monitored for continuous integration, continuous delivery, and continuous deployment.
o Specifically, MLOps provides reproducibility of ML pipelines, enabling more tightly-coupled collaboration across data teams,
reducing conflict with devops and IT, and accelerating release velocity.
 Risk reduction:
o Machine learning models often need regulatory scrutiny and drift-check, and MLOps enables greater transparency and faster
response to such requests and ensures greater compliance with an organization's or industry's policies.

MLOps with Databricks
 MLOps is a set of processes and automated steps to manage
code, data, and models. It combines DevOps, DataOps, and
ModelOps.
 Databricks recommends creating separate environments for
the different stages of ML code and model development with
clearly defined transitions between stages.
o Development
o Staging
o Production

MLOps with Databricks
 Development Stage
o The focus of the development stage is experimentation.
o Data scientists develop features and models and run experiments to optimize model performance. The output of the
development process is ML pipeline code that can include feature computation, model training, inference, and monitoring.
 Staging Stage
o The focus of this stage is testing the ML pipeline code to ensure it is ready for production. All of the ML pipeline code is
tested in this stage, including code for model training as well as feature engineering pipelines, inference code, and so on.
o ML engineers create a CI pipeline to implement the unit and integration tests run in this stage. The output of the staging
process is a release branch that triggers the CI/CD system to start the production stage.
 Production Stage
o ML engineers own the production environment where ML pipelines are deployed and executed.
o These pipelines trigger model training, validate and deploy new model versions, publish predictions to downstream tables or
applications, and monitor the entire process to avoid performance degradation and instability.

Why implement MLOps with Databricks?
 A unified platform
 Scalibilty and performance
 MLflow Integration
 Simplified ML workflow
 Automated deployment
 Improved Model Training

What are the components of MLOps?

Exploratory Data Analysis - EDA
 Exploratory data analysis (EDA) includes methods for exploring data sets to summarize their main characteristics and
identify any problems with the data.
 EDA can also influence which algorithms you choose to apply for training ML models.
 Databricks has built-in analysis and visualization tools in both Databricks SQL and in Databricks Runtime.

Data Preparations and Feature Engineering
 Data preparation:
o Cleaning and formatting data: This includes tasks such as handling missing values or outliers, ensuring data is in the
correct format, and removing unneeded columns.
o Preprocessing data: This includes tasks like numerical transformations, aggregating data, encoding text or image data,
and creating new features.
o Combining data: This includes tasks like joining tables or merging datasets.
o Storing the data: For storing the data we use Delta Lake, as Delta Lake is the optimized storage layer that provides the
foundation for tables in a lakehouse on Databricks. Delta Lake is open source software that extends Parquet data files
with a file-based transaction log for ACID transactions and scalable metadata handling.
 Feature Engineering
o It is the process of converting raw data into features that can be used to develop machine learning models.
o For ML applications, Databricks Feature Store helps your team discover and re-use features, track feature lineage, and
publish features to online stores for real-time serving and automatic lookup.

Model Training and Tuning
 Once you’re done with feature engineering, the next step is to train and tune your model and select the right algorithm or
architecture so that your model can achieve the desired accuracy.
 Use popular open source libraries such as scikit-learn and hyperopt to train and improve model performance.
 As a simpler alternative, use automated machine learning tools such as AutoML to automatically perform trial runs and create
reviewable and deployable code.

Model serving and deployment
 Databricks Model Serving provides a unified interface to deploy, govern, and query AI models.
 Each model you serve is available as a REST API that you can integrate into your web or client application.
 Why use Model serving?
o Deploy and query any models
o Securely customize models with your private data
o Govern and monitor models
o Reduce cost with optimized inference and fast scaling
o Bring reliability and security to Modell serving

References
 https://www.databricks.com/

Databricks for MLOps Presentation (AI/ML)

More Related Content

Similar to Databricks for MLOps Presentation (AI/ML)

Similar to Databricks for MLOps Presentation (AI/ML) (20)

More from Knoldus Inc.

More from Knoldus Inc. (20)

Recently uploaded

Recently uploaded (20)

Databricks for MLOps Presentation (AI/ML)