New Connected Data Workflows, Projects for Shared Context, and More Updates from Liminal
We’re excited to introduce a new set of features and enhancements designed to expand how teams securely use AI across their workflows! Over the past several weeks, we’ve introduced major improvements across user experience, model capabilities, and platform extensibility, while further strengthening security, compliance, and governance controls.
MCP Server Connections – Now Generally Available
Liminal continues to expand how organizations can securely leverage their data within AI workflows.
Users can access accurate, verifiable, and source-attributed results, with outputs re-ranked to prioritize the most current and relevant information across connected data sources.
Building on this foundation, MCP-powered integrations extend these capabilities beyond data access by enabling AI to securely connect to external systems and execute tasks across multi-step workflows.
This allows teams to move from retrieving and analyzing information to taking action within the systems where that work already lives.
As a result, teams can:
- Pull context directly from the systems they rely on
- Combine data into meaningful, multi-step workflows
- Take action across tools without breaking security or control boundaries
- Move from insight to execution without switching between systems
To help users get started, we have produced two onboarding guides:
Projects: A New Organizational Layer for AI Work
Projects, now rolling out, introduce a new way to structure and manage AI-driven work.
Instead of isolated conversations, teams can now group related threads under shared context, enabling more continuity, collaboration, and reuse.
With Projects, users can:
- Organize work by initiative, customer, or use case
- Maintain context across multiple interactions
- Collaborate more effectively across shared threads and workflows
- Lay the groundwork for multi-step, multi-thread workflows
This enables greater continuity and shared context across AI-driven work.
We have also created two onboarding guides to help users get started with Projects:
Introducing the Liminal Mobile App
Liminal is now available on mobile, giving users secure access to the platform beyond the desktop environment.
With the release of our mobile app, users can stay connected to their work and continue workflows from anywhere.
With mobile, teams can:
- Continue conversations and workflows from anywhere
- Quickly generate insights, summaries, or outputs in real time
- Stay connected to critical work without sacrificing security or governance
Whether it’s preparing for a meeting, responding to a customer, or reviewing outputs, AI is now accessible wherever work happens.
Faster, More Intuitive AI Interactions
We’ve significantly improved how users interact with Liminal day-to-day:
- Voice Dictation – Speak prompts directly into the platform
- Prompt Iteration Controls – Refine outputs with full control over settings
- Prompt Templates – Standardize common workflows and accelerate onboarding
These updates make it easier to move faster, experiment, and operationalize AI usage across teams.
Expanded Model Ecosystem
We continue to expand model support to give teams flexibility and performance across use cases:
- Added support for Claude Opus 4.8 and additional BYOM options
- Gemini 3.5 Flash now available as the default Google model
- Ongoing improvements to model selection, performance, and discoverability
This ensures users can choose the right model for each task without compromising governance or control.
File, Detection, and Platform Enhancements
We’ve also delivered a range of improvements across core platform capabilities:
- Expanded file support, including source code and structured logs
- Improved OCR, rendering, and upload reliability
- Enhanced detection accuracy across edge cases and multi-entity inputs
- Stronger system observability, logging, and error transparency
These updates continue to strengthen Liminal’s foundation in security, compliance, and enterprise-grade reliability.
Looking Ahead
We’re entering a new phase of the platform, focused on connecting AI to real work across systems, teams, and processes. This includes continued expansion of integrations and workflow capabilities, deeper orchestration across Projects, enhancements to the chat experience and complex output handling, expanded detection and policy enforcement, and much more.
See These New Features in Action
Check out everything that’s new in the Liminal Platform by scheduling a demo today!