April 2025 Summaries

The GenAI Platform has introduced a set of major updates, now in public preview, aimed at improving transparency, control, and performance visibility for users. These updates include enhanced agent development features and improved data sourcing through knowledge base citations, which provide clear source attribution for retrieved information. The platform also offers agent insights for performance monitoring, allowing teams to refine their AI agents for speed, efficiency, and cost-effectiveness. Additionally, agent versioning provides control over the evolution of AI agents, enabling teams to iterate with confidence and ensure continuous improvements without unexpected setbacks.

DigitalOcean has launched Managed Caching for Valkey, a fully managed in-memory key-value datastore designed to replace their previous Managed Caching service. This new service offers enhanced tools and is fully compatible with Valkey 8.0 and Redis 7.2.4 as a database. It provides benefits such as fast and efficient data storage, reduced database load, improved response times, and optimized resource usage. Existing customers with Managed Caching environments will have an end-of-availability (EOA) date of April 30, after which new clusters cannot be created, but existing clusters will remain operational. DigitalOcean Managed Caching for Valkey offers a range of features including automated failover, end-to-end security, high scalability, observability and monitoring, unique developer tools, and predictable pricing with flat rates and monthly caps starting at $15/month.

DigitalOcean is announcing its signing of DORA Addendums with eligible customers and has completed an audit of its environment by Schellman & Company to ensure compliance with the Digital Operational Resilience Act (DORA) requirements. The company recognizes the need to address DORA obligations for ICT third-party service providers, including cloud service providers, and is making the DORA Addendum available to select customers upon request. With a clean audit report, DigitalOcean has demonstrated its commitment to risk management, security controls, operational resilience, incident management, and outsourcing. The company's move aims to enable financial entities in the European Union and ICT third-party service providers to deploy on DigitalOcean with trust.

DigitalOcean has introduced a new product documentation chatbot designed to help developers quickly find accurate answers to their pressing product questions sourced directly from DigitalOcean's official documentation. The chatbot allows users to ask natural-language questions and get instant, documentation-backed responses, helping them stay focused on building instead of searching. It provides instant responses pulled directly from the official docs, uses natural language to interact in an intuitive way, has code snippets and examples at your fingertips, trusted information, seamless integration within the DigitalOcean documentation site, and is built specifically to make the workflow smoother and development process faster.

The GenAI Platform is being updated with new features that aim to make it easier to build, manage, and improve knowledge bases behind AI agents. These updates include web crawling capabilities, allowing users to crawl publicly available websites and index content into their knowledge base, reducing the need for manual data collection. Additionally, enhancements have been made to the Retrieval-Augmented Generation (RAG) system, which can help AI agents deliver faster, more accurate, and more context-aware responses from richer, better-organized data sources. The platform now also features Claude 3.7 Sonnet, a powerful reasoning-focused model that is ideal for agents requiring strong problem-solving, advanced comprehension, and trusted responses. Overall, these updates aim to elevate the capabilities of AI-driven projects by providing real-time web crawling, enhanced RAG capabilities, and access to advanced models like Claude 3.7 Sonnet.

The DigitalOcean GenAI Platform has introduced several new features and updates, including support for OpenAI models, improved web crawling for knowledge bases, and sensitive data detection guardrails. The platform now supports GPT-4o, o1, and o3-mini models, allowing users to integrate these advanced models into their AI agents using their own OpenAI API keys. Additionally, the GenAI Platform offers a range of tutorials, videos, and guides to help developers build and deploy AI applications, including real-time AI agents with serverless functions, knowledge graphs, and retrieval-augmented generation systems. The platform also provides features such as function calling, agent customization, and guardrails to facilitate the creation of secure, context-aware applications. With these updates, DigitalOcean aims to simplify AI development for all users, regardless of their experience level.

You can now bring your own OpenAI API key to use models like GPT-4o, o1, and o3-mini with GenAI agents on DigitalOcean, providing more flexibility when choosing the right model for your use case. The GenAI Platform supports industry-leading models from OpenAI, including generalist models like GPT-4o and reasoning models like o3-mini, allowing users to choose based on quality and capabilities. This integration enables developers to build AI-powered applications with a streamlined path to production-ready AI, without infrastructure headaches or re-architecting, using powerful models without setting up their own infrastructure, deploying agents with just a few clicks, connecting data to models easily, and getting predictable pricing and easy monitoring.

DigitalOcean has made several enhancements to its cloud storage solutions, including the general availability of per-bucket access keys with mixed permissions support and public preview of Spaces access logs. These updates aim to simplify storage management for customers handling large-scale deployments, automated backups, and security-driven workflows. The new features enable automation through various tools such as the DigitalOcean API, Terraform Provider, doctl CLI, and DigitalOcean's Python library (PyDo), allowing users to grant varying permissions to buckets with a single access key. Additionally, Spaces access logs provide detailed records of read and write requests, capturing object paths, client IPs, and more, which can be used for usage analysis, security, and compliance purposes. These enhancements are designed to improve the overall security, scalability, and manageability of DigitalOcean Spaces.

Makeissah Robinson is the Senior Director of Customer Support at DigitalOcean, leading her global customer support and service team with a focus on customer success and continuous improvement. She values diversity in her work environment and appreciates the company's emphasis on creativity, innovation, and continuous learning. As part of a company driving industry innovation, Makeissah believes she is making a real impact while being part of a supportive and innovative team that prioritizes its customers' needs. DigitalOcean's mission to act like owners, drive action, and simplify cloud computing resonates with Makeissah, who sees herself as part of an ecosystem working together to create technology that blazes the way.

DigitalOcean has introduced the general availability of its Partner Network Connect, a cross-cloud connectivity solution that simplifies and secures networking between DigitalOcean and other cloud providers or on-premise data centers. This solution enables businesses to establish direct, private connections, bypassing the public internet, to improve security, minimize latency, and increase network performance. Key benefits include improved latency, enhanced security, cost optimization, and simplified management. Partner Network Connect is available in all DigitalOcean regions except for a few, and its pricing is simple and predictable, with no charges on egress data transfer. The solution empowers businesses to seamlessly link their multi-cloud and hybrid environments, providing better performance, security, and cost control.

DigitalOcean is expanding its offerings by introducing the AMD Instinct MI300X GPU, which provides significant benefits for AI and high-performance computing (HPC) workloads. This new GPU offers faster deployments, increased control, and enhanced capabilities for developers. The system also includes a zero-configuration setup that reduces deployment time from weeks to minutes, making it easier to deploy popular AI models. Additionally, the platform simplifies pricing structures and provides access to ROCm developer resources, leading software platforms, and networking infrastructure within an open ecosystem. The GPU is now available on DigitalOcean in single-tenant Bare Metal configurations, allowing customers to scale AI/ML workloads on-demand without overprovisioning resources or incurring unnecessary costs.