Pinned update

Deep Hack hackathon announcement; Stable Diffusion workshop Stable Diffusion workshop on 25 and 26 March # On 25 and 26 March, Amir Nagri will teach a two-day online workshop (theory and hands-on) on Stable Diffusion. more

MLOps 2022 conference will bring together Machine Learning (ML), data engineering and SRE and AI practitioners to better understand the art of implementing MLOps in organizations.

MLOps is based on similar principles of DevOps, which talks about Continuous Delivery (CD) of ML models in production. We have seen significant traction for MLOps in recent years. Companies have been built around organizing and optimizing MLOps. We will focus on the tools, frameworks, data governance, decision-making process and approach based on real-world experience that have led to success and war stories.

The topics that we will focus in this conference include:

How do you add value to your MLOps processes?
- What are the key performance indicators?
- How do you quantify the impact of MLOps in your organisation?
What is the value addition to the business?
- How do you measure the Return of Investment (RoI)?
- What is the thriving work culture in your team/company?
How to design cost-effective pipelines?
- Are you investing in making the models leaner - distillation, quantization, pruning etc?
- What is the comparison of your hardware and performance costs?
What are the common mistakes made and lessons learnt in implementing MLOps?
- How did you resolve and recover from the mistakes made?
- Are you meeting SLA requirements and doing inference cost analysis?
What are some of your favourite tools for implementing MLOps?
- What are the key factors in deciding on a specific framework?
- When do you choose to build your own tool?
As an experienced MLOps practitioner, what advice do you provide?
- What advice would you give for beginners who are starting on MLOps?
- How do you scale your existing MLOps infrastructure?

In-person conference: MLOps 2022 will be held on Friday, 11 November, at Microsoft Reactor, Lavelle Road, Bangalore. This is an in-person conference. Register to participate for the conference.

Code of Conduct: Hasgeek’s Code of Conduct applies to the MLOps 2022 conference - participants, speakers and sponsors.

COVID protocols and masking policy: In keeping with COVID protocols, the following is applicable to all participants:

Participants attending in person must keep their vaccination certificate handy. The venue will ask you to show your vaccination certificate as proof of being fully vaccinated.
Wearing masks is optional.

Contact information: For queries about the conference, contact Hasgeek at support@hasgeek.com or call (91)7676332020.

Purchase a subscription to support The Fifth Elephant’s community activities on hasgeek.com

Hosted by

The Fifth Elephant

The Fifth Elephant - known as one of the best data science and Machine Learning conference in Asia - has transitioned into a year-round forum for conversations about data and ML engineering; data science in production; data security and privacy practices. more

Supported by

Primary Sponsor

Scribble Data

Scribble Data builds feature stores for data science teams that are serious about putting models (ML, or even sub-ML) into production. The ability to systematically transform data is the single biggest determinant of how well these models do. Scribble Data streamlines the feature engineering proces… more

Promoted

Community Sponsor

Thoughtworks

Our technologists build bold, future-focused solutions for clients from a variety of industries. You could influence the digital strategy of a retail giant, create a new mobile application for a bank or redesign platforms using event sourcing and intelligent data pipelines. more

Venue Sponsor

Reactor, Bengaluru, India

Join Microsoft Reactor and learn from developers Whether you’re building your career or the next great idea, Microsoft Reactor connects you with the developers and startups that share your goals. Learn new skills, meet new peers, and find career mentorship. Virtual events are running around the clo… more

All submissions

Previous Next

This submission has been added to the schedule

Do the Right Thing: Fast Development of ML applications using Sub-ML Feature Stores

Submitted Oct 27, 2022

Speaker #

Venkata Pingali, Co-Founder & CEO, Scribble Data

Abstract #

Widespread adoption of machine learning (ML) in industry is still a challenge today due to resource constraints and RoI questions. Production ML approaches today require high skill, rely on large volumes of data, and have long delivery timelines. In this talk, we argue for Sub-ML - a class of ML simpler than traditional ML approaches, often designed to be used in decision support systems, and delivered under tight constraints. Sub-ML, also called as ML-at-reasonable-scale (MLRS) and Analytical ML, covers upto 80% of the ML usecases in an enterprise. Characterized by their speed in realizing business value and support for diverse use cases, Sub-ML applications still require guarantees of correctness, transparency, and auditability in the data transformation process. We draw on our experience in the fin-tech, ed-tech and e-commerce domains to lay out design choices for feature stores to enable Sub-ML, tradeoffs we made including constraining the problem space, bundling capabilities for fast development, and incorporating a data consumption layer.