Generative AI Journeys - Fireside Chat with NinjaTech AI | Amazon Web Services

Science & Technology

Introduction

In this fireside chat, Sam, the co-founder and chief technology officer of NinjaTech AI, discusses the exciting world of generative AI with Amazon Web Services (AWS). NinjaTech AI is a company that aims to make everyone more productive by automating time-consuming tasks using multi-agent person AI. Their mission is to give back one hour of productivity time per week to every information worker around the world.

Sam shares his role at NinjaTech AI, as well as his history in the field of AI. He has previously worked on AI system codesign efforts at Meta and Google, optimizing AI models and building custom training infrastructure. NinjaTech AI is Sam's third startup, and it was born out of the realization that professional workers spend a significant amount of time on tasks that can be automated and delegated to an AI agent. Their focus is on democratizing access to personal AI, which has previously been out of reach for most people.

When it comes to the technology behind NinjaTech AI, Sam highlights the importance of leveraging AWS's custom AI silicon, specifically the TRN and Inferentia. The challenges of GPU availability and cost efficiency led them to choose AWS as their CSP. By using AWS's on-demand availability and cost-effective custom silicon, NinjaTech AI is able to offer an affordable personal AI to their users. Sam also mentions the support they receive from AWS, especially the Neuron team, which allows them to utilize the latest models and scale their operations on custom silicon.

SageMaker, AWS's machine learning platform, has also played a pivotal role in NinjaTech AI's journey. They have used SageMaker to extend their training jobs, leveraging TRN and conducting on-demand loads. The bursty nature of their experiments requires quick access to resources, and SageMaker provides them with the flexibility and scalability they need. Sam emphasizes the importance of not spending time on undifferentiated heavy lifting, allowing them to focus on building AI agents and skills.

Looking towards the future, NinjaTech AI has ambitious plans. They aim to expand the number of AI agents they offer and continue innovating with generative AI. Education and exploring different use cases for AI are also key aspects they want to focus on. By harnessing the power of generative AI, they believe they can have a profound impact across various industries and verticals.

Keywords:

  • Generative AI
  • NinjaTech AI
  • Multi-agent person AI
  • Productivity
  • Information workers
  • Custom AI silicon
  • TRN
  • Inferentia
  • AWS
  • SageMaker
  • Democratizing access
  • Personal AI
  • Machine learning
  • Scalability
  • Use cases

FAQ:

  • Q: What is NinjaTech AI's mission?

    • A: NinjaTech AI aims to make everyone more productive by automating time-consuming tasks through their multi-agent person AI, giving back one hour of productivity time per week to every information worker.

  • Q: Why did NinjaTech AI choose to leverage AWS's custom AI silicon?

    • A: NinjaTech AI faced challenges with GPU availability and cost efficiency. AWS's custom AI silicon, particularly the TRN and Inferentia, provided them with on-demand availability, cost savings, and the ability to scale their operations.

  • Q: How has SageMaker helped NinjaTech AI in their journey?

    • A: SageMaker, AWS's machine learning platform, has been pivotal for NinjaTech AI. It has allowed them to extend their training jobs, conduct on-demand loads, and quickly experiment with different models. SageMaker's flexibility and scalability have been crucial for their bursty applications.

  • Q: What are NinjaTech AI's future plans with generative AI?

    • A: NinjaTech AI aims to expand the number of AI agents they offer and continue innovating with generative AI. They also prioritize educating users and exploring different use cases across various industries and verticals.