A claimant uploads the completed claim form plus photographs of the damage, and one card appears in the queue with claim type, incident date, policy number, claimed amount, and a note of which supporting documents arrived. Triage happens on facts rather than on whoever opened the email first.
The app embedded on this page is the Application Tracker Board, a real Taskade Genesis build you can move cards through right now. It is a triage board in every respect except the reading, so start there: click "Use this app" to clone it in about ten seconds, relabel the lanes for your claim stages, then add the intake form, the reading step, and the extraction below.
The build is one intake with a queue and a completeness check behind it. A Claims Project holds one card per claim and an Evidence Project holds one row per uploaded item, kept as connected records so a claim with nine photographs is still one claim. A published form takes claimant name, policy number, claim type, and File fields for the form and the evidence. The Convert File to Text step reads each document, which is what lets a handwritten claim form sit beside a typed repair quote. An Ask Agent step returns claimant name, policy number, incident date, incident location, claim type, description of loss, claimed amount, third parties named, and the documents referenced in the form. It then compares that reference list against what actually arrived and reports the gap, which is the single most useful output in claims handling.
The queue reads where the work is:
- Board view to move claims through Received, Complete, Assessing, and Settled
- Table view for the full claims register with amounts and dates
- Calendar view for incident dates and assessment deadlines
- Plus List, Mind Map, Gantt, and Org Chart across the 7 project views
Automations move claims without a coordinator. With Taskade automations, a complete claim assigns to the handler on rotation, an incomplete one emails the claimant a list of exactly what is missing, and a claim above a value threshold notifies a senior assessor immediately. Across 100+ bidirectional integrations, forms and Gmail pull claims in while email and Slack push acknowledgments out. The AI agents inside the app carry 34 built-in tools with persistent memory, so an agent can summarize a claim file for the handler before they open it.
The completeness check is the leverage point, so spend your setup time there. List what a complete pack looks like per claim type, in plain words, and let the app compare that list against what arrived. Handlers then start with claims they can actually work rather than opening files to discover a missing quote. Set the value threshold that escalates a claim before you launch, and keep it as a field so it can be moved without a rebuild. Run a fortnight of real claims through with the routing switched off, read the triage the app proposed, and only then let it assign.
Clone the app, invite your handlers, and the claims, the evidence, and the routing rules are yours.
Read the upload path in File Uploads in Your App, the pattern in Build a Document Intake App, and the branching options in Build a Form with Conditional Logic. Set routing with Automation Triggers, browse live builds in the Community Gallery, or explore the templates library. Pair this with Track Certificates of Insurance Before They Expire for the policy side, and Digitize Paper Intake Forms at the Front Desk when claims still arrive on paper at a counter.
Frequently Asked Questions
Does it assess or decide the claim?
No. It reads what the form states, checks the pack for completeness, and routes the card. Assessment is a human judgment with obligations attached, and the build should be described that way to everyone who uses it.
Can it read photographs of damage as well as the form?
It reads words. A photograph of damage is stored as evidence and attached to the claim, and any text visible in it can be read, but the assessment of what the photo shows stays with a person.
How does the completeness check work?
The extraction reads which documents the form says are enclosed, then compares that against the files actually uploaded. The gap becomes the chase list, and the chase email writes itself from it.
What if the claim form is handwritten?
It is read off the page like any scan, and ambiguous values come back blank and flagged rather than guessed. That is the behavior you want when the field is a claimed amount.
Can claimants see their claim status?
On Business plans and above you can publish a signed-in portal on your own domain where each claimant sees only their own claim. App sign-in through App Users ties each view to a named person.
Are the uploaded documents retained?
Yes. Each file attaches to its evidence row and lands in your workspace Media tab, so the handler works from the originals.
What are the size limits on evidence?
Photos and scans should stay under 20 MB each, the ceiling for reading words off an image. Typed PDFs and Word documents are read as text and bounded by your plan upload size instead.
Can a claimant add documents after they submit?
Yes. Keep the upload surface open and tie later submissions to the claim reference, so a requested document joins the existing claim rather than opening a second one. The chase email carries that same link.
