Updates

We're working hard behind-the-scenes with updates to our dashboard and bringing forward our next ingest pipeline. In the meantime, we made some quick updates to our branding. We've also added new social channels to help you keep up-to-date.

• @joinsedna on X (formerly Twitter)
• @sednaos on Instagram
• and finally, on LinkedIn

We look forward to seeing you there!

Welcome to the new Sedna website! We've built this to better showcase what we're working on and make it easier for teams to understand how Sedna works and what problems it solves.

What You'll Find

The site gives an overview of what we're building, who we're building it for, and how to get started. If you're curious about what we're working on or want to learn more about AI governance and monitoring, take a look around.

We've expanded our provider coverage to include additional AI providers, especially those gaining traction in different regions and industries.

You can now monitor these new providers alongside your existing AI tools with the same analytics and cost tracking you're used to. Conversations, usage patterns, and cost data all flow into the same unified dashboard, so you don't need to jump between different tools to understand your AI spend.

We've built out our rule engine to give you more control over AI interactions. You can now set up custom rules to detect, modify, or block requests and responses based on your organization's needs.

Pattern Matching & Redaction

Create rules to automatically detect and redact sensitive information before it reaches AI providers. Use pattern matching to identify specific data types—credentials, personal information, proprietary data—and redact them in real-time.

Custom Blocking & Alerts

Set up blocking rules based on custom criteria and configure alerts for specific patterns. Want to prevent certain types of requests? Block them. Need visibility when specific patterns appear? Set up alerts. You can also analyze sentiment in requests and responses to flag potentially problematic interactions.

Runs in the Background

All rule processing happens automatically with no changes needed to your team's workflow. Rules are evaluated in real-time as requests flow through Sedna, keeping your organization safe without slowing anyone down.

We've rebuilt major parts of our infrastructure to make Sedna more flexible, faster, and easier to deploy across different network configurations and enterprise environments. This was our largest development push to date—over 90,000 lines of code shipped this month.

Multiple Deployment Options

You can now self-host remote instances of our ingestion pipeline, making it easy to upstream traffic from your existing infrastructure to Sedna. The platform supports multiple ingestion methods—ICAP servers, remote instances, and local agents—so you can integrate with whatever network setup you're running.

Rebuilt Traffic Engine

Our traffic ingestion engine has been decoupled and rebuilt. It can now accept traffic from multiple sources simultaneously and process it all through the same pipeline. This architectural change makes the platform significantly faster and more adaptable to different enterprise network configurations and security policies.

Enhanced Analytics Database

We've introduced TimescaleDB and restructured our database to handle massive volumes of analytics and usage data. This powers the dashboard with real-time insights into AI usage patterns, costs, and security incidents—even at enterprise scale with millions of daily requests.

Templated Rules Engine

The rules engine is now a templating system instead of hard-coded patterns. You can define your own security and compliance policies—match patterns with regex, use AI-powered detection with custom prompts, or identify specific strings. The system adapts to your needs rather than forcing you into predefined rules.

We've overhauled our internal development infrastructure to move faster and ship more reliable code. This includes simplifying our tooling, improving our deployment pipelines, and streamlining how we manage the entire development environment.

Simplified Tooling

We've moved from Node.js to Bun for faster execution, integrated Docker deep into our development workflow, and leaned harder into Next.js and oRPC for more granular control. The result is faster builds, more reliable systems, and the ability to spin up our entire infrastructure with a single command.

Better Security Management

We've fortified our internal infrastructure to be resistent to dependency injection attacks, and similar vulnerabilities. These hardening measures eliminate entire classes of potential security issues and ensure more secure operation across development and production environments.

Why This Matters

These changes don't just make us faster internally—they make Sedna better for you. We can iterate on features more quickly, deploy fixes faster, and maintain higher quality standards across the platform. The infrastructure improvements directly translate to a more stable, reliable product.

We've expanded support for AI-powered code editors and coding assistants, giving development teams visibility into how their engineers are using AI coding tools.

Development Team Visibility

Engineering leaders can now see patterns in how their teams use coding agents—what kinds of code generation requests are most common, which projects generate the most AI assistance, and where the team might benefit from additional training or tooling.

Same Dashboard, More Context

Coding assistant usage shows up in the same dashboard as your other AI tools, so you can compare coding tool usage with general AI adoption across your organization. All the same filtering, cost tracking, and incident detection works exactly the same way.

We've added comprehensive authentication and team-based access controls. Organizations need to make sure only the right people can see sensitive AI conversations.

Granular Permissions

You can now manage who has access to your organization's AI usage data with granular permissions. Set up multiple teams, assign different permission levels, and ensure that sensitive conversation data is only visible to authorized users.

Why This Matters

For compliance teams, this means you can give auditors read-only access without exposing configuration settings. For engineering teams, you can let developers see their own AI usage while restricting access to executive conversations. Everyone gets the visibility they need without compromising security.

We've rebuilt our core infrastructure to handle enterprise-scale traffic. We needed to make sure Sedna could keep up with thousands of concurrent AI requests without slowing down or missing data.

What Changed

We've improved the speed and reliability of our ingest pipelines with significantly better performance and lower memory usage. Better handling for long-running connections and streaming responses means more consistent data capture.

You'll notice faster response times and the ability to handle much higher request volumes without any degradation in monitoring capabilities. Sedna stays out of the way while still capturing everything.

We launched our MVP with support for a few third-party AI providers. You can now monitor conversations in real-time, track incidents, and understand usage patterns across multiple providers through a unified dashboard.

Three Months in the Making

Development has been ongoing since April. We've been heads-down building the core infrastructure, parsing engines, and dashboard while having conversations with interested investors about the future of AI governance and monitoring.

What's Included

We capture full conversation context across all supported providers. You get visibility into who's using which AI tools, what they're asking, and how much it's costing—all in one place.

Dashboard Features

The initial release includes live-updating charts, incident tracking, team management, and detailed cost breakdowns. Everything updates in real-time as your team interacts with AI providers.