01
Searching by hand
Someone opens a dozen documents and several mailboxes to assemble one answer, then starts again for the next request.
TraceFlow AI
MindScript IT · Evidence Intelligence for Manufacturing
Turn scattered product documents into verified answers.
TraceFlow AI structures product information scattered across PDFs, emails, spreadsheets and supplier documents, connects every value to the evidence behind it, and shows what is missing, inconsistent or out of date — so a manufacturing team can answer a client request with something it can stand behind.
MindScript IT is a software company. TraceFlow AI is our first product, and the one we are proving now.
Interactive prototype · fictional demo data · nothing to install
- Structured fields
- Evidence kept
- Gap detection
- Human validation
Incoming client request
“Please provide the material composition, country of origin, certificates and sustainability information for AeroStep Runner, SKU S-042.”
Giulia Ferrari · Meridian Sport Retail
Requested fields
- Upper material compositionFull-grain leather 78%, recycled polyester 22%Verified
- Lining materialRecycled polyester 100%Verified
- Sole materialNatural rubberSources disagree
- Country of assemblyItaly — Civitanova MarcheVerified
- Leather country of originNot providedNo source
Evidence behind the first answer
supplier_declaration_leatherworks.pdfPage 3
The upper is made of full-grain leather (78%) and recycled polyester textile (22%), supplied under contract L-18.
Client response
Draft — 4 of 7 items still without accepted evidence
The problem
Your information already exists. It is scattered.
A client asks a question your company can answer. The answer sits across PDFs, emails, spreadsheets, certificates, technical documents and supplier files — owned by different people, in different formats, from different years.
02
Disconnected systems
The shared drive, the mailbox, the spreadsheets and the business software each hold part of the picture. None of them holds all of it.
03
No way to tell what is current
A certificate may have expired, a specification may have been superseded, two documents may disagree — and nothing points it out.
MindScript turns that search into a process: structured fields, the source behind each one, and a person who approves what goes out.
The product
A product by MindScript ITTraceFlowAI
Evidence Intelligence for Manufacturing
What TraceFlow AI does, and what it refuses to do.
TraceFlow AI helps manufacturing teams structure information scattered across PDFs, emails, spreadsheets and supplier documents, connect important data points to their original evidence, and identify missing, inconsistent or outdated information.
TraceFlow AI turns scattered product information into structured, traceable evidence that helps manufacturing teams respond to client requests faster and with greater confidence.
TraceFlow AI is an interactive prototype. The demo uses fictional data and simulated extraction.
About TraceFlow AIHow it is different
Structured fields
Answers live in named fields with a defined meaning, not in a chat reply that has to be rewritten each time.
Source relationships
Every value stays attached to the document, page or cell it came from, so an answer can be checked rather than trusted.
Gap detection
Missing, expired and contradictory information is surfaced as work to do, instead of quietly passing through.
Human validation
Nothing becomes an answer without a named person accepting it. New evidence always returns for review.
Reusable responses
Once a field is verified with its evidence, the next request for the same information is an assembly job, not a search.
Who it is for
Manufacturers with complex product documentation. Footwear manufacturing in the Marche region is our chosen starting point for validating the workflow — a deliberately focused entry point, not a claim of existing customers or completed pilots.
TraceFlow workflow
From a client request to an answer you can stand behind.
Client Request → Structure → Evidence → Verify → Identify Gaps → Respond
01
Client Request
A client asks for composition, origin, certificates or environmental information. The request becomes a checklist of fields to answer.
02
Structure
The relevant documents are collected and read into those fields. Source collection belongs here, not as a separate project.
03
Evidence
Each proposed value is shown with the document, the page or cell inside it, and the sentence that supports it.
04
Verify
A reviewer accepts, corrects or rejects each value, and chooses between sources that disagree.
05
Identify Gaps
What is missing, out of date or contradictory becomes a targeted follow-up. New evidence returns to human review before it counts.
06
Respond
A structured answer to the original request: values, sources, who approved them, and anything still outstanding.
Source collection happens inside Structure. Resolving a gap loops back through verification — it never shortcuts into the response.
Test the Workflow
Interactive prototype · fictional demo data
Why now
Product information is being asked for more often, and more precisely.
Digital Product Passports are creating a growing need for structured, accurate and traceable product information. TraceFlow focuses on the evidence layer behind that information.
Client requests arrive more often
Retailers and buyers ask for composition, origin, certificates and environmental information as a matter of routine, and expect the document behind each answer.
Requirements are becoming more specific
European product rules are moving toward structured, machine-readable product information rather than documents sent on request.
The evidence layer is the hard part
Publishing a record is straightforward once the underlying information is structured, sourced and approved. That underlying work is what TraceFlow addresses.
TraceFlow is not a passport publishing tool. Passport readiness is one downstream use of the same evidence — client requests are the current focus.
How we build
Understand → Prototype → Validate → Scale.
How TraceFlow AI gets built: with manufacturing teams, on their own requests, one step at a time.
01
Understand
We sit with the people who answer client requests today and follow one real request end to end, including the parts that live in nobody's system.
The workflow, described honestly
02
Prototype
We build the workflow as usable software quickly, with real screens and real rules, so it can be judged rather than imagined.
A working prototype
03
Validate
A team runs it on their own documents and their own requests, and we measure whether it actually saves them the search.
Evidence it is worth building
04
Scale
Only once it holds up do we widen the product — more fields, more users, more of the process — so it becomes part of the workflow rather than another tool beside it.
Part of the daily workflow
The company
One product, and a narrow place to prove it.
MindScript IT is a software company. We do not take on custom development or consulting projects: we build one product and make it work properly for a specific kind of team before widening it.
01
One product
TraceFlow AI is our pilot product. Everything we build serves it, rather than a portfolio of one-off projects.
02
One industry first
Manufacturers with complex product documentation, starting with footwear in the Marche region — a deliberately narrow entry point, not a claim of existing customers or completed pilots.
03
Proof before scale
We validate on real client requests with the people who answer them today, and extend the product only once that holds up.
TraceFlow AI is an interactive prototype in validation. There is no deployed product, no customer base and no completed pilot to report yet.
Longer-term applications
- Supplier information workflows
- Compliance and audit preparation
- Digital Product Passport readiness
Not available today. These build on the same evidence layer once client requests are working well.
Team
Technology Leadership
Two people, one operating principle: start from a real operational problem and validate the solution with the people who live it.
Adem Bouteraa
Founder & CEO
Responsibilities
Product vision, business development and workflow validation.
Jihed Chouchane
CTO — Chief Technology Officer
Chief Technology Officer
Responsibilities
Technology strategy and technical architecture, AI systems, data extraction and intelligent workflow development.
Founder vision
Build intelligent business systems that turn scattered information into reliable, structured workflows — helping companies spend less time searching for information and more time acting on it. MindScript starts with real operational problems, validates solutions with real users, and builds technology that becomes part of the company's workflow rather than another disconnected tool.
Contact
Talk to us about a pilot.
Tell us about the client requests that take longest to answer, or the product information your team keeps searching for. We will tell you honestly whether TraceFlow AI is a fit, and what a pilot would involve.
- Phone+39 350 885 4726
- LocationAncona, Italy
- Emailcontact@mindscript.me
What happens after your message
- 01Understand — We sit with the people who answer client requests today and follow one real request end to end, including the parts that live in nobody's system.
- 02Prototype — We build the workflow as usable software quickly, with real screens and real rules, so it can be judged rather than imagined.