Import and create projects with AI

Use AI when the project shape already exists somewhere

Project AI is useful when the work is already scattered across a plan, export, meeting transcript, or rough brief, and you want ALLO to turn it into a project draft with sections and project canvases. Instead of copying rows one by one, you provide the source, review the generated structure, then create the project when it looks right.

This is still a project creation flow, not a finished project guarantee. AI can group work, suggest sections, create canvases, apply tags, add dates where it can read them, and include canvas subtasks when the source has checklist-like detail.

You still decide whether the structure matches the team’s real workflow. If you want to understand the project model first, read Understand projects and Create a project.

Availability

If the import or AI option is missing, check whether you can create projects in the current workspace and whether AI features are available for your account or plan. Missing actions are usually permission, plan, or workspace-setting issues, not lost data.

Choose import or AI-assisted creation

Import an existing plan or source

From the project creation flow, choose the import option, then choose the source family that matches your material.

The exact accepted file types depend on the source you select.

If you import from Trello, Asana, or Monday.com:

1. Export the project from that service first.
2. Upload the exported file to ALLO.
3. Review the draft ALLO creates from exported work names, descriptions, lists, sections, dates, and checklist-like content.

Attachments from those tools do not become ALLO attachments automatically. If the original source includes important files, add them separately as file objects or attach them to the relevant canvas after the project exists.

For meeting notes, use the cleanest source you have. A transcript with speaker chatter, repeated timestamps, and half-finished sentences may produce noisy canvases. If the transcript is messy, paste the action items or a cleaned summary instead of the whole meeting record.

Create from a prompt or template

When you choose AI-assisted creation, start with a project description or choose a template. The best prompt gives ALLO the operating shape of the project: the goal, major phases, deliverables, roles or teams, important dates, and any known checklist items.

Weak prompt: Plan our launch.

Better prompt: Create a project for the Q3 partner launch. Sections: planning, creative, partner review, launch week, follow-up. Include canvases for launch brief, landing page copy, demo video, email review, legal approval, partner enablement, analytics check, and retrospective. Add Due dates during August where they are mentioned.

Templates are useful when the work follows a known pattern but still needs adaptation. Pick a template close to the job, review the text ALLO puts in the prompt area, adjust it for your team, then generate the draft. Do not leave template language in place if it does not match the project; generic canvases become noise once real work starts.

Review the draft before creating the project

AI-generated projects should be treated as drafts. Review the preview before you create the project.

Check the project name first, because it will appear in Projects, Search, Home, Activity, and shared links. Use a specific name that the team can recognize later.

Review the sections next. They should match how work moves, not how the source happened to be formatted. If a spreadsheet has columns named Owner, Due, and Notes, those should usually become canvas fields or details, not project sections. Strong sections are phases, states, deliverable groups, or review stages.

Review canvases for:

• duplicates
• vague titles
• missing action verbs
• generated canvases that are not real work

Pricing is weaker than Review pricing page. If ALLO creates too many tiny canvases from a meeting transcript, keep the canvases that someone can own and move supporting details into canvas notes, subtasks, or sub-canvases after creation. See Create a canvas in a project.

Review tags. AI may create tags from repeated words in the source, but repeated words are not always useful tags. Keep tags that help people filter by:

• priority
• team
• channel
• risk
• customer
• work type

Clean up project-level tags with Manage project tags, and use Manage canvas tags when the project canvases need their own organization.

Review dates. Imported dates can be missing, ambiguous, or copied from unrelated text. Confirm that Due dates are real commitments before the project becomes visible to the team.

Review complete sections before relying on Dashboard. AI can suggest a workflow, but your team decides which sections count as finished work. Mark sections such as Approved, Ready, Published, or Done complete only when canvases there should count toward Project progress and burndown.

Permissions, access, and storage

Creating a project with AI uses your workspace permissions. If you cannot create projects, add members, change tags, or edit certain fields in ordinary Projects, AI creation will not bypass those limits. After the project is created, invite the right people and check access through Manage project members.

Project membership and canvas access are related but not identical:

• A teammate may see the project but not a restricted canvas inside it.
• An external collaborator may need access to a specific canvas rather than the whole project.

Share the smallest useful scope, then confirm that collaborators can open the work they need with Troubleshoot project and canvas access if anything looks blocked.

Imported source files are used as temporary import sources. They do not become normal Files items that people can preview, download, thumbnail, or reuse later from Files.

If the original source needs to remain part of the workspace record, upload it separately to a canvas or Files after the project is created. Normal file uploads, canvas file objects, and retained workspace files follow Storage and quota.

What can go wrong

The file is blocked before import. The type may not be supported for that source, or the file may be too large. Choose a smaller file, export a cleaner format, or split the source into smaller imports.

If a PDF is mostly scanned images, ALLO may not read the structure well. Export to CSV, spreadsheet, JSON, Markdown, or plain text when the source tool supports it.

The imported project looks plausible but misses context. Check whether the source uses unclear column names, abbreviations, merged cells, hidden sheets, incomplete exports, or private links that ALLO cannot read.

It can also happen when meeting notes contain discussion but not decisions. Add missing context in the prompt or source before trying again.

The draft has canvases without owners, dates, or enough detail. AI should not guess accountability when the source does not say who owns the work. Add owners after creation, or include owner names in the source before import.

A Trello, Asana, or Monday.com import is incomplete. Exports often separate work records from attachments, comments, automations, private fields, or connected files. Treat the ALLO project as the working structure, then add important supporting material manually.

Recover and troubleshoot

If upload fails, check the file type and size, then try a smaller or cleaner export:

• For meeting notes, paste the action-item summary instead of uploading the full transcript.
• For spreadsheet data, remove empty rows, hidden helper sheets, old archive tabs, and columns that do not describe the project.

If AI generation takes too long or fails, start with less source material. A smaller plan with clear canvas names, phases, and dates is better than a giant export with every historical comment. You can also create a blank project, add the first canvases manually, and use Use project views to organize the work after it exists.

If the project is created but looks wrong, fix the structure before inviting the wider team:

• rename sections
• remove duplicate canvases
• add useful tags
• make sure important briefs, files, notes, or deliverables live on the right canvases

If the wrong project was created and nobody has started using it, archive or delete it according to your workspace rules and create a cleaner version.

If you still cannot import, create, open, or recover the project, gather the source type, file name, approximate file size, workspace name, and what happened, then contact ALLO Support through Contact support.

After the project is created

Open the project and do one pass as the project owner. Confirm:

• name
• sections
• dates
• tags
• complete-section settings
• first canvases

Add members with Manage project members only after the structure is understandable. Use canvases where the team needs working space for notes, files, designs, decisions, or review material.

Keep the project small until the team starts using it. AI is good at drafting structure, but project quality comes from the review that follows: clear canvases, real owners, useful views, honest progress settings, and source material that people can find later.