AI Platform Highlights: Amazon Nova, Solo.io MCP, Kong Gateway, and More

Software companies continuously add AI features to their platforms, while AI firms release new models and capabilities, making it challenging to keep up; this overview shares key AI updates developers should know.

Anthropic Announces Claude Participation in Education Projects

Anthropic introduced a new curriculum that includes a "learning mode" encouraging students to solve problems together with Claude rather than having Claude provide answers alone. For example, Claude may ask "What evidence supports your conclusion?" to help students better understand their work, and it also emphasizes core concepts while offering templates for research papers, study guides, and outlines.

Project website: https://www.anthropic.com/news/projects

Amazon Creates New Site to Explore Amazon Nova Foundation Models

Amazon launched nova.amazon.com , providing information on each Amazon Nova foundation model such as Nova Micro, Nova Pro, and Nova Canvas. Each model has its own page with details, benchmarks, demos, and getting‑started resources.

In addition, Amazon announced Amazon Nova Act, a new model that can perform actions in a web browser. The company released a research preview of the model SDK so developers can start experimenting.

Rohit Prasad, Senior Vice President of Amazon Generative AI, said the site puts the power of Amazon’s cutting‑edge intelligence into the hands of developers and enthusiasts, enabling rapid testing of ideas before scaling them on Amazon Bedrock.

Solo.io Launches MCP Gateway to Address AI Agent Tool Sprawl

Solo.io announced the MCP Gateway, a model‑context‑protocol gateway for the cloud‑native API gateway formerly known as Gloo. MCP, an open protocol developed by Anthropic, standardizes how applications connect data sources and tools to large language models. As MCP adoption grows, AI agents face tool overload and challenges in discovering, connecting, and securing diverse MCP‑compatible tools.

The MCP Gateway aims to solve this by letting developers discover, protect, and federate multiple MCP tools and services into a virtualized MCP server, providing a single access point regardless of how many tools exist in the AI agent ecosystem.

Kong AI Gateway Updates to Reduce LLM Hallucinations and Protect Sensitive Data

Kong updated its AI Gateway (version 3.10) with new features, including the RAG injector that automatically queries vector databases and injects relevant data to reduce LLM hallucinations, ensuring results are enhanced with known knowledge sources.

Placing a vector database behind the Kong AI Gateway also improves security and lets developers focus on building rather than mitigating hallucinations, boosting productivity.

Env0 Announces Cloud Analyst

Env0 introduced Cloud Analyst, an AI agent that provides insights into an organization’s entire cloud infrastructure. Users can ask about infrastructure status, trends, and key metrics, explore historical data filtered by project or environment, discover optimization opportunities, and create interactive dashboards. The service will launch in Q2 as part of the env0 platform and is currently on a waiting list.

CodeSignal Adds AI Skill Assessments

CodeSignal added AI skill assessments to measure workplace AI competencies. The assessments include AI literacy, rapid engineering, and AI researcher evaluations. CEO and co‑founder Tigran Sloyan noted that, according to the World Economic Forum, over 40% of workers will need to reskill in the next three years due to AI and automation, and the AI Collection offers a precise, scalable way for organizations to assess and develop these abilities.

Zencoder Releases New AI Coding and Unit Test Agents

Zencoder launched new AI agents that integrate with more than 20 development tools, including VS Code, JetBrains IDEs, JIRA, GitHub, GitLab, and Sentry. The agents can perform complex multi‑file refactoring and merging, automatically test and optimize output. CEO Andrew Filev said the agents act as force multipliers, handling routine coding so engineers can focus on innovation and more complex software engineering tasks.