AUTODESK EXPERT SPEAKER · AUTODESK UNIVERSITY 2026

Stop adapting your business to the software.We adapt the software to your business.

Name: Vault on Autopilot: AI-Driven PDM Automation Across Lifecycle, Revisions, and ERP
Start: 2026-09-17
End: 2026-09-17
Location: Autodesk University 2026

My purpose is to serve by solving problems. I'm Jose Luis Baquero Rivera, Senior Consultant, AI Solutions: I build AI agents and automations that live inside Vault, Inventor, AutoCAD and Forma. Technology is the means; the real work is removing friction from engineering teams' day-to-day.

MESSAGE ME ON WHATSAPP

VAULT · INVENTOR · AUTOCAD · FUSION · FORMA

>angle bracket 240 × 90, height 120, thickness 10, with 4 holes Ø12

PART ANGLE BRACKETNATIVE FEATURES, EDITABLE IN THE MODEL TREEREV A

RECREATION OF THE AI DESIGN ASSISTANT · SAMPLE DATA

THE DIFFERENCE

Auditing an assembly before release: 20 minutes, or one.

The same task, side by side. A reconstruction of a real Vault Professional workflow.

BY HAND · VAULT CLIENT 00:00

01Find the assembly and open every subcomponent
02Check lifecycle state part by part
03Verify revisions against the BOM in Excel
04Spot files still in Work In Progress
05Confirm required properties on every card
06Write down findings and draft the report email

Repeated for every release. Every step is a chance for error.

VAULTPILOT · ONE COMMAND 00:00

>

✓142 components traversed with your credentials
✓3 files still in WIP, with owner and date
✓1 revision out of sync with the ERP BOM
✓Report ready. No writes without your confirmation

Same permissions, same Vault. What changes is who does the clicking.

ILLUSTRATIVE RECONSTRUCTION OF A REAL WORKFLOW · TIMES MEASURED ON CLIENT

THE PROOF

What I've built

Four cases, each with its real constraint and the decision that solved it.

01FLAGSHIP PRODUCT

VaultPilot

The AI brain for Autodesk Vault Professional — in partnership with Avant Leap

CONTEXT: Engineering teams with thousands of files in Vault Professional, where every release demands repetitive manual checks.
CONSTRAINT: An AI can't write freely into a PDM: it must respect permissions, schema and traceability, or it never reaches production.
DECISION: A native agent that operates with the user's own credentials, with dry-run and confirmation before every write, validated against the live Vault schema.
RESULT: 28 tools in production across three surfaces — Vault Client, AutoCAD 2026 and Inventor 2026 — and the architecture selected for Autodesk University 2026.

VaultPilot — AI agent for Autodesk Vault

02AI IN INVENTOR

AI Design Assistant

From Chat to CAD — tell Inventor what to build, and it builds it

CONTEXT: Repetitive modeling tasks an engineer can describe in one sentence but takes far longer to build by hand.
CONSTRAINT: The generated model must be parametric and editable inside the active assembly — not dead geometry.
DECISION: A three-layer agent architecture — Inventor API, MCP and LLM — on a .NET add-in with access to Content Center and templates.
RESULT: Parametric parts and assemblies generated from natural language, with a pattern replicable to any Inventor design workflow.

AI Design Assistant in Autodesk Inventor

03DATA AUTOMATION

Content Center + Piping

Thousands of families and styles, managed from Excel and code

CONTEXT: Migration and standardization of conduit and piping libraries across Inventor versions.
CONSTRAINT: Managing Content Center by hand doesn't scale, and a mass error in the families is hard to roll back.
DECISION: A suite that exports full libraries to Excel, batch-creates Tube & Pipe styles and edits families at the XML level with backups and logs, using the already-authenticated Vault Explorer session.
RESULT: Libraries migrated and standardized without rebuilding them by hand, with a history of every change.

04AUTODESK FORMA

Forma as a CDE

Information governance at infrastructure scale

CONTEXT: An airport infrastructure program with multiple companies working on the same information.
CONSTRAINT: A CDE must comply with ISO 19650 while still allowing configuration changes without breaking live projects.
DECISION: A master hub template with permissions and inheritance defined in Docs, ISO 19650-aligned metadata and a test hub ahead of production.
RESULT: Review workflows, issues, forms, model coordination and cost management under a single source of truth, replicable to new projects.

MORE SCOPE EXAMPLES

Vault migrationThree vaults, 51 libraries and 100 GB moved from Vault 2022 to Vault 2026.BUILT

Automated skid designAutomated structural design of industrial skids inside Inventor.BUILT

iLogic rules and familiesiPart tables, Content Center families and iLogic rules to generate product configurations.BUILT

DocBridgeA bridge between BIM 360, Forma and SharePoint: it migrates with metadata and version history, and publishes in under 60 seconds.PRODUCT

Vault linked to the ERPERP data available in real time while filling in each file's properties.DESIGNED

P&ID drawings and quotes in AutoCADAutomatic numbering, smart blocks, bill of materials to Excel and ERP integration.DESIGNED

Terrain in Civil 3D with DynamoAutomated terrain workflows, planned in six phases.DESIGNED

Budget, cost and construction planningA catalog of automations on Autodesk Platform Services.DESIGNED

BUILT: DEVELOPED OR IMPLEMENTED · DESIGNED: SCOPE DEFINED, NOT YET BUILT

HOW WE WORK

How we work

We start with a single process. You see the result with your own data and decide whether we continue.

01
Diagnosis
You tell us where hours are being lost. We tell you frankly what can be automated, what can't, and how much effort it takes.
YOU GETThe list of what is worth doing, in order.
02
Pilot
A single process running on your data, so you see the result before committing to more.
YOU GETThat process solved and in use.
03
Implementation
We connect it to what you already use and support your team until they adopt it.
YOU GETThe solution inside your operation.
04
Support and evolution
We keep adjusting the solution as your operation changes.
YOU GETSoftware that does not fall behind.

WHO'S BEHIND IT

You work directly with me

AUTODESK EXPERT · AU 2026 SPEAKER · DUAL MBA

No middlemen, no forms: the person who diagnoses the problem, builds the solution and answers the WhatsApp is the same one.

I've spent more than a decade solving problems in the Autodesk ecosystem — first implementing and selling it, now extending it with artificial intelligence. I founded asistIA to build AI agents, and today I build AI solutions for Autodesk products as a Senior Consultant at Avant Leap.

I work from Colombia for LATAM and globally remote, in Spanish and English.

+200% growth in revenue and prospect connections, reported by a client.

PUBLIC LINKEDIN RECOMMENDATION

Stop adapting your business to the software.We adapt the software to your business.

Autodesk selected this architecture for its flagship technical event.

Vault on Autopilot: AI-Driven PDM Automation Across Lifecycle, Revisions, and ERP

Auditing an assembly before release: 20 minutes, or one.

What I've built

VaultPilot

AI Design Assistant

Content Center + Piping

Forma as a CDE

Solving problems. That's the service.

How we work

Diagnosis

Pilot

Implementation

Support and evolution

You work directly with me

Let's talk about your case