Follow the conversation, inspect the workflow board and read the work behind an individual step in its thread.
AlphaAgent Studio
Build. Direct. Run.
Studio is the working environment your teams use to build specialist agents and teams, direct open-ended work and run governed workflows, with plans, progress, evidence and outputs together in one place.
Inside Studio
Studio in action.
Move from configuring specialists to directing live work and running proven workflows without switching products.
Slide 1 of 5
Work in Studio
Two ways to work.
Use the same specialists, intelligence and controls for open-ended work and governed workflows. Choose the working shape that fits the objective.
Open-ended work
Follow the evidence.
When the objective evolves as new questions and evidence emerge, talk naturally with a specialist in Studio Chat. Studio works out the grounding and coordination the task needs, with Governor intervention checkpoints at meaningful decisions. Keep the evidence and outputs together, and turn a useful path into a draft workflow.
Set the direction.
Guide the specialist and respond at Governor intervention checkpoints as the objective develops.
Trace the work.
Use the Configuration Inspector to see the agent version, intelligence and connections in play. Review citations and web sources, then replay the path through the Thought Map.
Keep the outputs.
Browse and download workspace files. Reuse persistent datasets across the conversation.
Governed workflows
Run within bounds.
When the outcome and controls are defined, build a visual sequence of agent tasks, document analysis, conditions and human handoffs. Steps can pass full datasets forward and return structured outputs. Review and activate a version, then run it on demand or describe its schedule in plain language.
Adapt to cases.
Use conditions, retries and timeouts without forcing every case through one rigid path.
Hold decisions.
Pause for approval, review, guidance or a stop decision wherever judgement is required.
Preserve every run.
Review the exact workflow version, each step, decision, dataset and artefact from one complete record.
Work in depth
Answers become assets.
Studio brings source material, current information and connected data into one working context, then returns results your team can inspect and use again.
- 01 / Inputs
Bring the evidence.
Attach documents and images. Draw on Intelligence Graphs, enterprise data, approved external tools and the live web.
- 02 / Analysis
Run the analysis.
Studio reads the inputs, retrieves connected data and runs Python or shell analysis while progress streams live. Agents call your existing Java libraries for pricing, risk, credit or actuarial calculations in an isolated execution environment, without rewriting the models your teams trust.
- 03 / Outputs
Keep the result.
Explore full datasets as grids and charts. Keep generated reports and files in the conversation workspace, ready to inspect, reuse or download.
The conversation stays readable. The full data, working files and outputs remain available across turns and when you return.
Review and download the PDFs, presentations, documents, spreadsheets and diagrams an agent produces.
Configure in Studio
Assemble the capability.
Start with the outcome, then configure the specialists, reusable intelligence and coordination required to carry the work. Choose the tools and authority each specialist needs.
One configured capability
- 01 / Define
Specialist agents
Describe the specialist you need. Studio drafts its system prompt for you to review. Each edit creates a new version. You can also pull a specialist shared through your Organisation Library.
- 02 / Equip
Reusable intelligence
Pin the specialist to chosen Intelligence Graph and connector versions. Connect SQL databases, REST APIs, AWS sources and approved tools through the Model Context Protocol (MCP). Live web search and browsing are also built in.
- 03 / Coordinate
A specialist team
Bring agents together under a supervisor that interprets the objective, divides the work and routes each part to the right specialist.
Ready for work.
Direct the capability through live work. Turn a proven path into a governed workflow without rebuilding the intelligence, connections, team or controls underneath.
Bring existing Claude skills, memories and MCP servers into Studio. Build on what your teams already use.
Live work
Stay in command.
Studio keeps the objective, plan, progress, evidence and output together. People can manage live and completed work, direct it where judgement matters and improve it without supervising every step.
- The work stays visible.
- Long analyses, workflow runs and Intelligence Graph builds continue on the server if you leave the page or close the tab. Activity cards show the current step and progress from anywhere in Studio, and notifications take you back to work that has finished or needs attention.
- Change without surprises.
- Agents, workflows, connectors and Intelligence Graphs keep immutable versions. Agents stay pinned to the graph and connector versions chosen for them until someone changes them. Conversations and completed runs record what was used.
- People set direction.
- Approve an investigation plan before it runs, add guidance while work is underway and use workflow handoffs for approval, review, guidance or a stop decision.
- Keep work aligned.
- Governor supervises every agent by default. It checks each step against the objective and grounded facts, steers work back when it drifts and sends the agent back to its sources when needed. Its guidance stays on the step’s card.
Organisation Library
Reuse what works.
Let more teams build on proven work. Share a selected version through the Organisation Library. Authorised teams pull a complete copy into their own Studio and choose whether to adopt later source updates.
- 01
Prove the work.
Build and refine an agent, workflow or Intelligence Graph in Studio.
- 02
Share a version.
Choose the version to share. A workflow carries its agents, intelligence, connectors and environments with it.
- 03
Make it yours.
Pull a connected copy as a draft. Review it, then choose when to activate it. Its origin and source version remain recorded.
Built in Studio
Programmatic access.
Build agents, specialist teams and workflows in Studio. Run them there, on a schedule, or from your own systems. In a governed deployment, everyone uses the same active workflow version. Organisations controls the keys that let business systems start it.
- In Studio
- People direct agents and teams, run workflows and review the work in one place.
- Via API
- An administrator issues a key for one deployment and its permitted workflows. Your systems start runs, follow progress and collect outputs in your AWS account. Approval steps wait in Studio’s Inbox. Only a person can release them.
Deployment boundary
Your AWS. Private access.
AlphaAgent runs in AWS accounts you own and connects directly to your AWS Transit Gateway, with zero access from the public internet. Your agents, workflows, intelligence and results stay inside your deployment boundary.
Customer-controlled boundary
Your AWS accounts
Only network ranges you approve can reach AlphaAgent, with sign-in through your Entra ID and MFA. Your network, DNS, identity and firewall policies control access.
AlphaAgent Studio
AlphaAgent OS
AlphaAgent Organisations
Private connections
- Amazon Bedrock
- Your data
- Your systems
Amazon Bedrock and AWS services use private VPC endpoints. Your prompts, research and results never cross the public internet. Agents can reach only approved network ranges and ports, with read-only SQL access.
Customer work stays here
- Agents and workflows
- Specialist intelligence
- Conversations and documents
- Data and credentials
- Results and evidence
Outbound onlySigned licence heartbeats and usage counts
Outbound traffic follows your firewall, allow-list, inspection and logging policies.
Architecture diagrams and network checklist
AlphaAgent OS
One foundation underneath.
Every agent, team, conversation and workflow inherits the same intelligence, execution, authority and evidence as work moves through Studio.
Explore AlphaAgent OSReusable intelligence
Intelligence Graphs and Memory give specialists trusted expertise and continuing context that can be reused across work.
Coordinated execution
Kernel, Flows and Bridge run agents and tools together, preserve state and connect work to approved enterprise systems.
Built-in control
Authority, Governor and Lens define decision rights, keep agent behaviour within bounds and make every run inspectable.
Why Studio
Engineer every use case.Transform the business outcome.
Studio makes sophisticated multi-agent systems straightforward to configure. The complete operating environment is already built.
Other platforms
Build the system.
Models
Frameworks
Cloud
You engineer the rest.
Agents
Architecture, prompts and expertise.
Controls
Data access, memory and permissions.
Assurance
Continuous testing and evidence.
Experience
A working user interface.
Low code helps with assistive agents. Core work needs repeatability across users.
Months of engineering.
Before work can run. Then much of it repeats.
AlphaAgent
Transformation at scale.
Studio
Build and run AI teams.
OS
Power and govern work.
Organisations
Manage access. Share proven work.
Console
Manage usage and billing.
Make expertise reusable.
Turn institutional knowledge into Intelligence Graphs. Equip highly specialised agents.
Scale what works.
Turn proven agent work into governed workflows. Run them at scale.
Stay in command.
Set authority. Test performance. Supervise behaviour. Inspect evidence. Approve what matters.
Configure your capability. Reuse the complete system.
Start in Studio