Thesis question01
Can margins remain resilient?
Pressure-test Meridian’s supply chain and order book.
NPNina Patelowner
Axolect Beta
Watch the Run. Steer the Research.
Axolect is the shared AI research workspace for investment teams. Watch agents work, guide research during a run, and review proposed changes together.
Beta coming soon
Shared context for teams. Human review for proposed changes.
A looping product tour of three investment analysts sharing one live research canvas. Nina runs a Claude Code agent on a fictional source pack. She invites Priya, who brings her own Codex agent onto the same canvas. Alex opens Priya’s agent from his laptop, watches the run, and sends guidance mid-run. Priya hands the agent off to Alex with its sources and run log attached. Alex then approves the proposed evidence, and the canvas updates for all three teammates.
01 / 08Scripted · fictional data
Built for investment teams working together on AI research.
Portfolio ManagersResearch AnalystsRisk TeamsInvestment DesksResearch Groups
The Problem
AI research needs a shared workspace.
Investment research involves analysts, portfolio managers, and risk teams. Agent work should be visible to the people shaping the research.
Agent work stays in personal chats
One analyst sees the prompt and the agent response. Everyone else receives a summary later, without the working context behind it.
Guidance arrives after the run
A portfolio manager or risk teammate may spot a missing assumption while research is underway. A finished answer leaves that guidance for the next run.
Proposed changes need a shared review
When findings arrive separately from the thesis, the team must reconstruct which evidence supports a change and who agreed to make it.
The Research Workspace
Context, agent work, and review on one canvas.
Axolect brings the research context and the people reviewing it together. Follow the run, guide the next step, and decide which changes belong on the canvas.
Research context the team can follow
Connect theses, hypotheses, and evidence on the canvas. Keep the research question and its supporting context alongside agent work.
Thesis · Hypothesis · Evidence
Agent activity visible as it happens
Follow an agent run through its output and status. Shared visibility gives the team a place to assess how the research is developing.
Run output · Status · Shared context
Guidance while research is underway
Work on the canvas together and send guidance to an active agent run. Add context when it matters to the research in progress.
Team presence · Live sync · Mid-run guidance
Review with a record of decisions
Review agent proposals before they change the canvas. Revisit the proposal and the recorded decision when the team needs to understand a change.
Proposals · Approvals · Recorded decisions
The Shared Workflow
Keep the team involved as research unfolds.
Shared Research Context
Keep theses, hypotheses, evidence, and agent work on the same canvas so teammates can follow the research in context.
Visible Agent Runs
Watch agent output and run status in the workspace. See work develop while there is still time to guide it.
Guidance During a Run
Authorized teammates can send guidance to an active run, adding context or redirecting the research as it progresses.
Review Before Changes
Agents propose canvas changes for review. Approval decisions are recorded so the team can revisit what changed and why.
How It Works
From shared context to reviewed changes.
Set the research context
Put the thesis, hypotheses, and evidence on a shared canvas. Give the agent a research question grounded in that context.
Watch the agent work
Start a run and follow its output and status. Teammates can see the research develop in the same workspace.
Guide the active run
Send guidance while the agent is running. Clarify the question, add an assumption, or ask it to focus on a particular line of research.
Review proposed changes
Review the proposed nodes or relationships together. Authorized reviewers approve or reject changes, and the decision is recorded.
Who It Is For
One workspace for the people shaping research.
Investment Teams
For analysts, portfolio managers, and risk teams.
Investment desks that need to follow agent work, contribute research context, and review proposed changes in a shared workspace.
- Workspace access with defined roles
- Shared canvases and visible agent runs
- Proposal review with recorded decisions
Research Groups
Work through research with invited teammates.
Small research groups that want a common place to develop hypotheses, guide agent work, and evaluate proposed findings together.
- Collaborate within a shared workspace
- Keep hypotheses and evidence connected
- Follow runs and review proposed changes
Where We Are Going
Built with investment teams, through beta.
Evaluate the current workflow and help shape what comes next.
Today Shared AI Research
A shared canvas for agent runs, guidance during a run, and proposal review with recorded decisions.
Planned Run Hand-off
We plan to let teammates take responsibility for a live run with its working context intact. Run hand-off is not yet available.
On the Roadmap Stable Release
Beta feedback will inform the stable release, including which research workflows need refinement. Scope and timing are not yet announced.
Planned work is forward-looking. Run hand-off and the stable release are not yet available; scope and timing may change.
Beta Questions
Frequently Asked Questions
Learn how shared runs, guidance, and proposal review work in the beta.
What is Axolect Beta?
Axolect is the shared AI research workspace for investment teams. Watch agents work, guide research during a run, and review proposed changes together.
Who is the beta for?
The beta is for investment research teams, including analysts, portfolio managers, risk teams, and investment desks. Invited research groups can evaluate the shared workflow too.
Can I guide an agent while it is working?
Yes. Authorized teammates can send guidance during an active run to clarify the research question or add context. The guidance is recorded with the run.
How are proposed changes reviewed?
Agents produce structured proposals for canvas changes. Authorized reviewers can inspect, approve, or reject them under workspace review rules. Decisions are recorded alongside the research.
What is planned after the beta?
Run hand-off is planned and is not yet available. Beta feedback will inform the stable release. Pricing, release scope, and timing have not been announced.
Explore the Beta
Bring your team into the research.
Beta access is coming soon. Follow shared agent work, guide active runs, and review proposed changes together.
Beta coming soon