Log In
Sign Up
MLOps
Developing a Scalable MLOps Cloud Infrastructure for Edge Deployment of Vision Machine Learning Models Contract
Job Description:
Join the Factbird Team and Take Flight into a World of Data Revolution!
At Factbird, we’re on a mission to transform data collection and production monitoring, empowering organizations worldwide with the transformative power of data.
We’re on the lookout for one (or maybe even two!) talented students eager to embark on a thrilling Master’s thesis journey with us. If you’re studying Computer Science, AI, Software Technology, or related fields at DTU, KU, ITU, or a similar institution, this opportunity could be tailor-made for you.
Here are the exciting details:
️ Start Date: Ideal kickoff in early 2024 (January or February), depending on your semester schedule.
⏳ Duration: The project will align with the scope of your Master’s thesis.
Project title: Developing a Scalable MLOps Cloud Infrastructure for Edge Deployment of Vision Machine Learning Models
Executive Summary:
The rapid advancement of machine learning (ML) has propelled its adoption across various industries, particularly in the realm of computer vision. However, the effective management and deployment of ML models, especially for edge devices, present significant challenges. This thesis project aims to address these challenges by developing a scalable MLOps cloud infrastructure for training, deploying, and evaluating vision ML models in an edge deployment setting.
Problem Statement:
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Project Objectives:
The primary objectives of this thesis project are to:
The primary objectives of this thesis project are to:
- Develop an MVP MLOps cloud infrastructure that seamlessly integrates edge deployment capabilities
- Design a unified framework for managing the entire ML lifecycle, including training, deployment, evaluation, and monitoring
- Enable BYOM functionality, allowing customers to easily integrate their proprietary ML models into the platform
- Optimize the infrastructure for real-time inference on edge devices
Technical Approach:
The proposed MLOps cloud infrastructure will leverage Amazon Web Services (AWS) services, including Amazon Greengrass, Amazon SageMaker, and Amazon Lambda, to provide a comprehensive solution for managing and deploying vision ML models. The infrastructure will be designed to efficiently handle multiple tenant workloads, ensuring scalability and availability for edge deployment.
Key Deliverables:
- A detailed MLOps cloud infrastructure architecture that incorporates edge deployment capabilities
- Automated workflows for training, deploying, evaluating, and monitoring models
- Real-time inference capabilities for edge devices
Expected Outcomes:
The successful completion of this thesis project will result in:
- A robust and scalable MLOps cloud infrastructure for edge deployment of vision ML models
- Streamlined ML lifecycle management processes
- Real-time inference capabilities on edge devices
- Enhanced productivity and efficiency for ML model development and deployment in edge applications
Impact:
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
A little about us
Factbird is a game-changing end-to-end manufacturing intelligence solution that simplifies data gathering and analysis for all manufacturers. With this solution at your fingertips, you can finally take charge of your manufacturing efficiency and redefine the boundaries of success. Factbird, in business since 2015, is now used in over 25 countries worldwide and trusted by 200+ manufacturers daily, with offices in Denmark and the US. After a recent funding round, we are currently scaling significantly across the globe.
Our story is a narrative of ceaseless innovation, collaborative growth, and a resolute commitment to our customers. We’ve continually evolved, developing our product with our clients’ feedback, establishing features and solutions that not only restructure industry operations but cement our position as a trailblazer in manufacturing technology.
We’re scaling significantly across the globe after a recent funding round, and we want you to be a part of our exciting journey! Join the Factbird Team, and you’ll enjoy
- Flexible working hours.
- Skills development and growth opportunities.
- A warm and structured onboarding
- You choose your gear to work with what inspires you.
- An international, diverse, and inclusive culture with colleagues from 17 different nationalities!
Ready to embark on an incredible journey and shape the future of data collection and production monitoring?
Don’t miss this exciting opportunity – join us now by submitting your application! Plus, feel free to share this awesome chance with anyone who’d be a perfect fit!
Our interview process is a breeze – it starts with a chat with our recruiter, followed by one technical interview, and culminates with a talk with our leadership team. Got any questions or need more details? Reach out to Khristel Marie Aparicio at khma@factbird.com
Join the Factbird Team and Take Flight into a World of Data Revolution!
At Factbird, we’re on a mission to transform data collection and production monitoring, empowering organizations worldwide with the transformative power of data.
We’re on the lookout for one (or maybe even two!) talented students eager to embark on a thrilling Master’s thesis journey with us. If you’re studying Computer Science, AI, Software Technology, or related fields at DTU, KU, ITU, or a similar institution, this opportunity could be tailor-made for you.
Here are the exciting details:
️ Start Date: Ideal kickoff in early 2024 (January or February), depending on your semester schedule.
⏳ Duration: The project will align with the scope of your Master’s thesis.
Project title: Developing a Scalable MLOps Cloud Infrastructure for Edge Deployment of Vision Machine Learning Models
Executive Summary:
The rapid advancement of machine learning (ML) has propelled its adoption across various industries, particularly in the realm of computer vision. However, the effective management and deployment of ML models, especially for edge devices, present significant challenges. This thesis project aims to address these challenges by developing a scalable MLOps cloud infrastructure for training, deploying, and evaluating vision ML models in an edge deployment setting.
Problem Statement:
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Project Objectives:
The primary objectives of this thesis project are to:
The primary objectives of this thesis project are to:
- Develop an MVP MLOps cloud infrastructure that seamlessly integrates edge deployment capabilities
- Design a unified framework for managing the entire ML lifecycle, including training, deployment, evaluation, and monitoring
- Enable BYOM functionality, allowing customers to easily integrate their proprietary ML models into the platform
- Optimize the infrastructure for real-time inference on edge devices
Technical Approach:
The proposed MLOps cloud infrastructure will leverage Amazon Web Services (AWS) services, including Amazon Greengrass, Amazon SageMaker, and Amazon Lambda, to provide a comprehensive solution for managing and deploying vision ML models. The infrastructure will be designed to efficiently handle multiple tenant workloads, ensuring scalability and availability for edge deployment.
Key Deliverables:
- A detailed MLOps cloud infrastructure architecture that incorporates edge deployment capabilities
- Automated workflows for training, deploying, evaluating, and monitoring models
- Real-time inference capabilities for edge devices
Expected Outcomes:
The successful completion of this thesis project will result in:
- A robust and scalable MLOps cloud infrastructure for edge deployment of vision ML models
- Streamlined ML lifecycle management processes
- Real-time inference capabilities on edge devices
- Enhanced productivity and efficiency for ML model development and deployment in edge applications
Impact:
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
A little about us
Factbird is a game-changing end-to-end manufacturing intelligence solution that simplifies data gathering and analysis for all manufacturers. With this solution at your fingertips, you can finally take charge of your manufacturing efficiency and redefine the boundaries of success. Factbird, in business since 2015, is now used in over 25 countries worldwide and trusted by 200+ manufacturers daily, with offices in Denmark and the US. After a recent funding round, we are currently scaling significantly across the globe.
Our story is a narrative of ceaseless innovation, collaborative growth, and a resolute commitment to our customers. We’ve continually evolved, developing our product with our clients’ feedback, establishing features and solutions that not only restructure industry operations but cement our position as a trailblazer in manufacturing technology.
We’re scaling significantly across the globe after a recent funding round, and we want you to be a part of our exciting journey! Join the Factbird Team, and you’ll enjoy
- Flexible working hours.
- Skills development and growth opportunities.
- A warm and structured onboarding
- You choose your gear to work with what inspires you.
- An international, diverse, and inclusive culture with colleagues from 17 different nationalities!
Ready to embark on an incredible journey and shape the future of data collection and production monitoring?
Don’t miss this exciting opportunity – join us now by submitting your application! Plus, feel free to share this awesome chance with anyone who’d be a perfect fit!
Our interview process is a breeze – it starts with a chat with our recruiter, followed by one technical interview, and culminates with a talk with our leadership team. Got any questions or need more details? Reach out to Khristel Marie Aparicio at khma@factbird.com
Executive Summary:
The rapid advancement of machine learning (ML) has propelled its adoption across various industries, particularly in the realm of computer vision. However, the effective management and deployment of ML models, especially for edge devices, present significant challenges. This thesis project aims to address these challenges by developing a scalable MLOps cloud infrastructure for training, deploying, and evaluating vision ML models in an edge deployment setting.
Problem Statement:
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Traditional MLOps methodologies often lack the flexibility and scalability required for edge deployment, making them less suitable for real-world applications. Current solutions often focus on specific aspects of the ML lifecycle, such as training or deployment, but fail to provide a unified and comprehensive approach. Additionally, the ability for customers to bring their own models (BYOM) further complicates the management and deployment process.
Project Objectives:
The primary objectives of this thesis project are to:
The primary objectives of this thesis project are to:
Technical Approach:
The proposed MLOps cloud infrastructure will leverage Amazon Web Services (AWS) services, including Amazon Greengrass, Amazon SageMaker, and Amazon Lambda, to provide a comprehensive solution for managing and deploying vision ML models. The infrastructure will be designed to efficiently handle multiple tenant workloads, ensuring scalability and availability for edge deployment.
Key Deliverables:
Expected Outcomes:
The successful completion of this thesis project will result in:
Impact:
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
The proposed MLOps cloud infrastructure will significantly impact the adoption and utilization of ML models in edge devices. By providing a streamlined and secure solution for model management and deployment, the infrastructure will enable businesses to harness the power of ML for real-time edge applications, such as anomaly detection, image recognition, and object tracking. Additionally, the BYOM functionality will empower businesses to leverage their own proprietary ML models without compromising their intellectual property.
A little about us
Factbird is a game-changing end-to-end manufacturing intelligence solution that simplifies data gathering and analysis for all manufacturers. With this solution at your fingertips, you can finally take charge of your manufacturing efficiency and redefine the boundaries of success. Factbird, in business since 2015, is now used in over 25 countries worldwide and trusted by 200+ manufacturers daily, with offices in Denmark and the US. After a recent funding round, we are currently scaling significantly across the globe.
Our story is a narrative of ceaseless innovation, collaborative growth, and a resolute commitment to our customers. We’ve continually evolved, developing our product with our clients’ feedback, establishing features and solutions that not only restructure industry operations but cement our position as a trailblazer in manufacturing technology.
We’re scaling significantly across the globe after a recent funding round, and we want you to be a part of our exciting journey! Join the Factbird Team, and you’ll enjoy
- Flexible working hours.
- Skills development and growth opportunities.
- A warm and structured onboarding
- You choose your gear to work with what inspires you.
- An international, diverse, and inclusive culture with colleagues from 17 different nationalities!
Ready to embark on an incredible journey and shape the future of data collection and production monitoring?
Don’t miss this exciting opportunity – join us now by submitting your application! Plus, feel free to share this awesome chance with anyone who’d be a perfect fit!
Our interview process is a breeze – it starts with a chat with our recruiter, followed by one technical interview, and culminates with a talk with our leadership team. Got any questions or need more details? Reach out to Khristel Marie Aparicio at khma@factbird.com
