Partner Success with AWS / Software & Internet / United Kingdom

December 2024
Untitled design - 1
Rebura

NeuralSpace Accelerates AI Model Training Speed by 96% in Migration to AWS with Rebura

Learn how AI startup NeuralSpace migrated to AWS to train its models at speed and to do so affordably.

96%

reduction in AI model training time

25%

cost saving over previous provider

2

FTE made available for more productive work

Overview

NeuralSpace, a London-based AI startup, had the same problem that many startups have: not enough time, not enough money, and too much to do. It needed to develop and train the AI models that powered its language AI applications—automatic translation of text and speech, automated subtitling, and automated AI dubbing of content—but these processes were taking too long. With 20–30 TB of data being used to train each model, it could take 3–6 months to train just one. And the company needed to train multiple models to develop its products. NeuralSpace knew that it needed to find a way to speed up model training that would fit within its limited budget. With the help of AWS Partner Rebura, NeuralSpace migrated to Amazon Web Services (AWS) to enable faster modeling and a crucial pivot in focus.

NeuralSpace

Opportunity | Struggling to Train AI Models Fast Enough

Founded in 2021, NeuralSpace had focused on developing natural language processing (NLP) AI tools. It was developing four core products on another cloud provider and managing and training its models manually across all four. However, this was time-consuming, expensive, and inefficient, and it hampered the company’s plans and its ability to scale. NeuralSpace needed to find a way to cost-effectively speed up its model training. The company was approached by AWS, which had noticed the work that the company was doing. “We were trying many things,” says Ayushman Dash, chief executive officer (CEO) and founder of NeuralSpace. “We even tried to build our own infrastructure to save money. And somehow the universe gave us the solution when someone from AWS got in touch with us and said, ‘We can solve this problem for you.’ They introduced us to Amazon SageMaker.”

NeuralSpace’s AWS account manager introduced the company to a few AWS Partners and NeuralSpace hit it off with Rebura, a London-headquartered specialist with additional offices in the Netherlands, Sweden, and Switzerland. By this time, NeuralSpace had realized that it needed to pivot its focus. Instead of developing four products in parallel, it wanted to focus on the offering that had the most market traction. The AI product is intended for users in creative and creative-adjacent roles and offers three tracks: StorytellerAI, an AI digital twin to help generate ideas, music, and scripts using the user’s own style; AI Dubbing, a way to provide fast, affordable AI-voiced dubbing for content; and AI Subtitling, which offers AI-powered subtitling that’s twice as fast as other options. “We’re here to augment human creativity using AI, and we believe in human-powered AI, not AI-powered humans,” says Dash. “We believe that AI will help human beings to be more creative but always retain the human at the center. That’s our ethos.”

kr_quotemark

We cut training time by up to 96 percent and our costs are down by about 25 percent. Amazon SageMaker makes it so easy.”

Ayushman Dash
Chief Executive Officer and Founder, NeuralSpace

Solution | Using Amazon SageMaker to Accelerate Training and Cut Costs

Rebura explained that it could build the infrastructure that NeuralSpace needed to use Amazon SageMaker to train its models in days instead of months. It was also able to explain how the startup could qualify for service credits that would make the migration and development of its new infrastructure affordable. Rebura and NeuralSpace worked together to create a list of the capabilities the startup wanted and to construct an AWS infrastructure that would affordably deliver what was needed. This is where the companies found that they were a great fit and worked well together. “I think the synergy was there from the start as we understood what NeuralSpace was trying to do and how our expertise in AI workloads could help them achieve that,” says Sam Samarasekera, UKI sales director at Rebura. “We were only founded 5 years before NeuralSpace, so we understand the challenges a startup faces. Every startup needs to speed up time to market and time to revenue, and it’s our role to build what they need to make that happen.”

NeuralSpace and Rebura each focused on its individual area of expertise—NeuralSpace developing its applications and Rebura building the infrastructure in what Samarasekera describes as “a very collaborative process.” At the heart of the system is Amazon SageMaker, a fully managed service that brings together a broad set of tools to enable high-performance, low-cost machine learning (ML) for any use case. With SageMaker, users can build, train, and deploy ML models at scale. That was a game-changer for NeuralSpace because it had previously manually managed model training, a task that required 1–2 staff devoted to the job.

Other key components of the infrastructure include Amazon Simple Storage Service (Amazon S3), an object storage service offering industry-leading scalability, data availability, security, and performance, and Amazon FSx for Lustre, a fully managed shared storage system. “FSx for Lustre is very important, because a major bottleneck in training AI models on large amounts of data is the speed of moving data from your data sources to your models,” says Dash. “FSx for Lustre overcomes that.” The company also uses Amazon Elastic Kubernetes Service (Amazon EKS), which allows it to efficiently run distributed training jobs using the latest Amazon Elastic Compute Cloud (Amazon EC2) GPU-powered instances, which Dash describes as “essential to my business.”

Outcome | AI Model Training Now 96% Faster and Costs Reduced by 25%

NeuralSpace has been transformed by the migration, which has enabled the company to pivot and focus on its most promising innovation. It simply could not do this before, because model training was too slow and too costly, holding back the company. “Working with Rebura was an amazing experience,” says Dash. “They’re incredibly professional, have deep technical knowledge, and the communication was perfect. We just had to tell them what we needed, and they made it happen. Our AWS account manager was also involved throughout the process—we felt really well supported.”

Amazon SageMaker has transformed the speed of training the AI models, making it possible to train a model in just 7 days instead of 6 months. “It costs quite a bit to train a model in 7 days,” says Dash. “You’re consuming a huge amount of resources to run through 20 or 30 TB of data. But, in the big picture, it actually cuts costs because we only pay for what we use when we use it. We spin it down when we’re not in use. We cut training time by up to 96 percent and our costs are down by about 25 percent. SageMaker makes it so easy.”

NeuralSpace is now 90 percent migrated to AWS after just 4 months and plans to migrate the remaining 10 percent in the near future. During the migration process, the company also got better insight into its operations, making compliance easier for its ISO 27001 security certification and for GDPR certification. It also used the migration as an opportunity to optimize its backend processes. “I wish we’d started out with Rebura and AWS when we launched,” says Dash. “It would have saved a lot of time and money and, as a startup, you always need more of both.”

About NeuralSpace

Founded in 2021, NeuralSpace is an AI-focused startup providing translation services. The company, with headquarters in London, UK, and offices in India, worked on multiple AI tools focused on natural language processing. It has pivoted to focus on using its language expertise to develop tools to assist those in the creative arts. It offers tools that help users to generate ideas, music, and images, all tailored to fit their style; create subtitle content in more than 100 languages; and produce high-quality dubs for content using AI voices.

About AWS Partner Rebura

Rebura is an AWS Advanced Service Partner that helps customers accelerate their technology transformation. Focused exclusively on AWS, the company uses its tech, data, and cloud expertise to help customers get the most out of theirs. From engineering and project management to 24x7 managed services and FinOps, Rebura have the skills and resources to support customers at any stage of their AWS journey.

AWS Services Used

Amazon S3

Amazon Simple Storage Service (Amazon S3) is an object storage service offering industry-leading scalability, data availability, security, and performance.

Learn more »

Amazon SageMaker

Bringing together widely-adopted AWS machine learning and analytics capabilities, Amazon SageMaker delivers an integrated experience for analytics and AI with unified access to all your data.

Learn more »

Amazon EKS

Amazon Elastic Kubernetes Service (Amazon EKS) is a fully managed Kubernetes service that enables you to run Kubernetes seamlessly in both AWS Cloud and on-premises data centers.

Learn more »

Amazon FSx for Lustre

Amazon FSx for Lustre provides fully managed shared storage with the scalability and performance of the popular Lustre file system.

Learn more »

More Software & Internet Success Stories

Showing results: 9-12
Total results: 149

no items found 

  • Software & Internet

    Palo Alto Networks Boosts 2,000 Developers’ Productivity Using AI Solutions from AWS, Anthropic, and Sourcegraph

    Palo Alto Networks, a leading cybersecurity company, sought to boost developer productivity using generative artificial intelligence (AI) technology. The goal was to create a custom solution that would enhance the speed and quality of coding while maintaining strict security standards. By leveraging Amazon Web Services (AWS), Claude 3.5 Sonnet and Claude 3 Haiku from AWS Partner Anthropic, and Cody from AWS Partner Sourcegraph, Palo Alto Networks developed a secure AI tool for generating, optimizing, and troubleshooting code. Within three months, Palo Alto Networks onboarded 2,000 developers and increased productivity up to 40 percent, with an average of 25 percent. This custom AI solution has empowered both senior and junior developers, and the company expects further improvements in code quality and efficiency.

    2024
  • Software & Internet

    IBM Reduces the Co-Selling Lifecycle by 90% and Boosts Sales Opportunities with AWS by 117% with ACE CRM Integration Using Labra Platform

    IBM, a global technology enterprise, wanted to simplify the process of creating and sharing AWS co-selling opportunities from Salesforce. IBM deployed an ACE CRM integration from AWS Partner Labra, a provider of software as a service (SaaS) solutions. The integration helps IBM sales and marketing teams move campaign responses and sales opportunities from within Salesforce directly into ACE. With Labra’s co-sell automation, IBM has cut co-sell time by 90 percent, increased co-sell opportunities by 117 percent, increased revenue, and created a custom integration that streamlines marketing nurture tools.

    2024
  • Software & Internet

    Starburst Accelerates AWS Co-Selling with Tackle ACE CRM Integration

    Starburst, which provides an open data lakehouse platform for global customers, sought to reduce the manual effort required to collaborate with its strategic cloud partner, AWS. Starburst uses Salesforce as its CRM system and needed a solution to replicate relevant opportunity data to the APN Customer Engagements (ACE) pipeline manager. Starburst worked with AWS Partner Tackle and implemented the Tackle ACE CRM integration, which allows Starburst to enter, manage, and monitor sales activity to enable sophisticated co-sell activity within ACE, the AWS Partner Network (APN) sales collaboration tool. As a result, Starburst cut opportunity sharing time by up to 50 percent, reduced opportunity rejections, enhanced opportunity data quality, and enabled teams to focus on other joint go-to-market (GTM) opportunities.
     

    2024
  • Software & Internet

    ERIN’s Cloud Transformation Cuts Costs, Promotes Growth, and Enables New Features

    An innovative employee referral platform operating with a team of 25 people, ERIN faced challenges keeping up with rapid growth. A collaboration with AWS Partner SourceFuse improved ERIN’s cloud setup to use its resources more effectively and eliminate unnecessary services. The outcome was transformative: ERIN cut hosting costs by 40 percent, found the flexibility it needed to meet demand, and advanced its feature roadmap. The solution helped ERIN achieve near triple-digit growth in revenue year over year, broaden its reach, and deliver new offerings to customers.

    2024
1 38

Get Started

Organizations of all sizes across all industries are transforming their businesses and delivering on their missions every day using AWS. Contact our experts and start your own AWS journey today.