We’re excited to announce our partnership with Amazon Web Services (AWS). Our team has optimized Refact.ai to run on Amazon Elastic Compute Cloud (Amazon EC2) instances, which provide secure and resizable compute capacity. This collaboration establishes Refact.ai as the first enterprise AI assistant for coding available on AWS Inferentia2 instances via AWS Marketplace.
Adapting the StarCoder2 model to run on Inferentia2 instances using AWS Neuron, Refact.ai delivers 1.5x higher performance per dollar and 20 percent faster speed than comparable GPU setups. This solution ensures enterprise data security with SOC 2 compliance, allowing companies to deploy Refact.ai within their AWS environments. By leveraging AWS Inferentia2, Refact.ai offers a cost-effective and reliable AI assistant for business solutions, ensuring organizations keep full control over their sensitive data.
Refact.ai easily integrates into AWS environments, supporting development in IDEs like VS Code and JetBrains with compatibility across 25+ programming languages. Leveraging models like Llama3/8b and Starcoder2, optimized for AWS Neuron, our solution provides advanced code auto-completion powered by Retrieval-Augmented Generation (RAG) and chat capabilities deeply aware of the company’s codebase. These tools empower developers to get more accurate code suggestions, generate precise code, efficiently find and fix bugs, review code, and more.
By leveraging AWS’s latest Inf2 instances, powered by high-performance, low-latency chips, Refact.ai achieves up to 1.5x better performance per dollar and a 20 percent speed increase relative to other comparable instances, making it the most cost-effective solution for enterprise development teams.
“This approach delivers immediate benefits for large teams, offering minimal latency and high performance,” highlights Refact.ai’s senior developer, Kirill Starkov.
The ease of deployment through the AWS Marketplace enables companies to integrate Refact.ai into their workflows with minimal setup and immediate results, making it a competitive option among top cloud providers.
Refact.ai offers deployment options tailored to the security and compliance needs of enterprise organizations. It can be deployed on an Amazon EC2 instance within a customer’s AWS environment or on-premises, ensuring sensitive data, code, and intellectual property remain fully under the organization’s control.
All processes run locally, avoiding reliance on external services or third-party platforms, and the deployment model is SOC 2 compliant, combining strict security standards with the scalability and cost-efficiency of AWS infrastructure. For organizations with specific regulatory requirements, on-premise deployments ensure complete data ownership while leveraging these deployment options make Refact.ai a flexible solution for cloud-based deployment in industries that prioritize control and compliance. This empowers enterprises to enhance development workflows and efficiency without compromising security or operational integrity, demonstrating the value of secure cloud deployment.
“Given the high cost and limited availability of GPUs, partnering with AWS allowed us to overcome a critical challenge for scaling in this industry—access to high-performing and cost-effective compute resources. With AWS Inferentia2, we’ve addressed this need by providing enterprises with a secure and reliable solution that ensures exceptional performance while significantly reducing inference costs,” says Oleg Kiyashko, Cofounder of Refact.ai.
Looking to the future, we are developing an Autonomous AI Agent built to manage complex engineering tasks end-to-end. Unlike traditional tools, this Agent will act as a digital twin for developers, integrating seamlessly with tools that developers usually use to access the broader context beyond the codebase. Equipped with built-in memory, it will continuously learn from developer interactions, build a comprehensive knowledge base, and evolve over time.
At Refact.ai, we are committed to empowering developers with advanced AI tools. If you’re ready to integrate Refact.ai into your workflows through AWS Marketplace, let’s explore how our AI solution can enhance your software engineering teams while ensuring the highest levels of security and control.
Our collaboration with AWS marks a pivotal moment in shaping the future of AI-driven software creation. We’re excited to bring our developer productivity tools to enterprises worldwide and can’t wait to see how your teams improve overall development efficiency with Refact.ai!
Stay (fine-)tuned as we continue to build the future of programming