.jpg)
Migration mapping that ran 6 to 9 months now finishes in hours, with the write-up produced alongside.
DeepShift Migration AI
AI Exchange Hackathon
Fraud Detection & DMCA Takedown AI
Copyright claims are classified against past cases in seconds instead of days, with an analyst deciding.
Orchestrated AI workflow that classifies copyright violations against historical case data, with human review.
What the agent does
Scrapes a suspected infringing page, classifies it against past takedown decisions, and routes the call to an analyst with its reasoning.
- Brand
- BÆRSkin Tactical
- Contributed by
- MACH Alliance (Hub operator)
- Industry
- Apparel & Footwear, Retail
- Audience
- B2B
01
The problem
Why this exists
Brand teams sit on years of DMCA takedown history in tools like Trello and Monday.com, but every new suspected violation triggers manual triage: a person opens the URL, compares it against past cases by memory, decides whether it crosses the line, and starts the takedown workflow.
Read more
The volume is increasing and the institutional memory is locked inside whoever has been doing it longest. The brand needs to scale the work without losing the audit trail or handing decisions to a model with no oversight.
02
Outcomes
What this demonstrates
- A submitted URL is classified against historical DMCA cases in seconds, not days.
- Classification confidence and the case-history evidence land in Slack, where a human analyst approves, overrides, or escalates.
- Analyst decisions feed back into the case graph, so the model gets sharper as the brand processes more cases.
Read more
- The audit trail of who decided what, against what evidence, is preserved automatically. Useful for legal and for compliance.
- The workflow is model-agnostic. The same orchestration runs against GPT-4, Claude Sonnet, or a future model without rewiring the pipeline.
03
How it works
Architecture
A URL enters the workflow through a Slack submission or a webhook. The Conscia DX Engine orchestrates the steps: a page scraper pulls the live content from the submitted URL, the DX Graph is queried for similar historical cases (vectorized case descriptions plus structured fields like brand, country, and decision), and an LLM, OpenAI GPT-4 or Claude Sonnet depending on configuration, classifies the case against the precedent and emits a confidence score with cited evidence.
Read more
The result returns to Slack as a structured message with the classification, the cited historical cases, and approve / override / escalate actions. When the analyst chooses an action, that decision flows back through a webhook into DX Graph, so the system's memory of "what counts as a violation for this brand" stays current with the people doing the work.
Historical case data was imported from Trello and Monday.com exports as part of project setup. PDF process documentation was ingested at the same time so the model has access to the same policy briefs the human analysts work from. The whole orchestration (scrape, retrieve, classify, route, capture feedback) runs as a single DX Engine workflow with usage-based pricing measured in API calls.
Tech stack
Conscia DX EngineConscia DX GraphOpenAI GPT-4Claude SonnetTrello / Monday.com (case data)SlackVercel
Next step
Take the code
This one is published in full. Read it, run it, fork it. Nothing here is behind a conversation.
Request an intro
This one is the MACH Alliance’s own work. Ask about speeding up your claims triage, and the people who built it will answer.
Not ready to talk to anyone? MACH Insider is the Alliance mailing list. New solutions, program openings, and the research behind them, once a month.
Related solutions
Wholesale buyers place and change orders in plain language, and see the price they will actually pay.
SynapseAI B2B Wholesale
AI Exchange HackathonEvery promoted item is checked against the source of truth, with problems surfacing in minutes, not weeks.
Wegmans Marketplace Validation Agents
AI Exchange Hackathon
Verification
Public code
The implementation is published in full, so anything on this page can be checked against the repository. Anything the repository does not show is not established here.
Status: Reference prototypeAdded May 26, 2026A listing is an editorial decision, not a MACH Alliance certification.
Built something worth showing?
Put your agentic solution in front of the enterprise leaders who use the hub to decide what to build next. Submissions are open to MACH Alliance Members, Collaborators, or Ambassadors.