Hackathon Portal
AI Tinkerers - Barcelona
Team

IFX

Project Concept

IFX — Instance Finance

IFX is a verification layer for financed productive assets.

Traditional asset finance treats collateral as a static object. In reality, financed assets are living systems: they operate, consume energy, generate output, issue invoices, receive payments, trigger incidents and change risk over time.

IFX turns those operational and financial signals into reproducible proof bundles.

The first demo asset is a financed GPU cluster. IFX verifies whether the cluster exists, runs workloads, consumes plausible energy, generates invoices, receives payments, services debt and respects the agreed waterfall.

For each period, IFX produces a machine-readable evidence bundle containing:

  • Operational signals
  • Invoice and payment reconciliation
  • Debt-service calculations
  • DSCR and LTV metrics
  • Audit flags
  • Integrity score
  • Cryptographic proof hashes

The goal is not to issue securities or settle payments. The MVP demonstrates the missing evidence layer for future asset-backed finance: continuous verification of whether productive assets are real, active, revenue-generating and compliant with their financing rules.

IFX is useful for lenders, insurers, auditors and asset owners who need more than periodic reporting. It provides continuous, auditable evidence of productive cashflows.

Entry

Status: Not Started

Team Roster

You must be registered for the event to view the team message board.

Jose Vargas Team Lead RSVP Approved

Vocational Teacher in Industrial Automation and Robotics at Institut El Palau
I am a vocational teacher in industrial automation and robotics, with hands-on experience in PLCs, SCADA, industrial robots, drives, machine vision and fabrication. Alongside teaching, I am building a technical project around AI agents that can control software, machines and real-world workflows more reliably. My main interest is the interface between AI and physical systems: automation, manufacturing, simulation, machine control and practical prototyping. I like turning unusual technical ideas into testable systems.
AI agents for controlling software, machines and industrial workflows; structured UI/DOM representations; multimodal reasoning with visual/interface primitives; AI-based sensing for millifluidic reactors using DSP, communications algorithms, data acquisition, PINNs, simulation and digital twins. I would like to meet people working on agents, embodied AI, embedded systems, industrial sensing, robotics and simulation.
I am building text/DOM interfaces for Windows and Linux to help AI agents control software more reliably. The current line of work is a JSON-DOM representation of user interfaces: structured snapshots of UI elements, state and possible actions. The goal is to generate training and evaluation data for software-control agents, instead of relying only on screenshots or pixels. This is part of a broader effort toward AI-controlled industrial workflows and small automated production cells.