TATIANA MARTINS@tatianatenorio798652

Problem. Brazil has 5,570 municipalities and 27 states, each with its own tax rules, and the tax reform (Complementary Law 214/2025) runs until 2033. Companies must review prices and product catalogs, courts face a growing volume of retroactive calculations, and a single missing or wrong document can disqualify a multimillion tender. Users. Companies, accounting offices, law firms, court accounting teams, judicial experts and public-tender bidders. Solution. Vyson Calc & Negócios is a multi-agent platform with four engines: a tax module that audits invoices and ERP catalogs in batch (XML, PDF, CSV, Excel); a judicial engine that recalculates court cases in batch with control number, report and opinion; a pricing optimizer that protects the margin after the reform; and a tender automator that checks documents to reduce errors. AI on AMD. Files come in batch (up to 1,500 per ZIP) and are audited in parallel. Evolus document extractors and rule agents check the data in a fixed sequence. Gemma 4 (Google DeepMind), served on AMD GPU through the AMD Token Factory, writes each opinion in one call, with personal data masked before the call. A citation lock rejects laws or numbers not found in official sources. AI explains, the official table decides, the human signs. Measured (live test, Oct 17, 2026). Search across 1,533 official sources in 0.005 s; AI opinion in about 12 s; 7 of 7 sample invoices audited with opinion in 17 to 24 s. Validation. A tax review reduced a paying client's tax burden from 20.8% to 6.5% (-69%), within the law. Two public tenders won and awarded, totaling R$ 2.36 million. The founder was part of the Court of Justice of Alagoas team that returned 7,600 cases. Business. SaaS by volume, pay per batch on AMD, success fee on awarded tenders and a partner network. Demo: https://tatianabot-dashboard.vercel.app/fiscal

Vyson FiscalRAG is a retrieval-augmented generation container built for AMD AI League Match 3. Pointed at a folder of mixed enterprise documents (PDF datasheets with superseded revisions, DOCX roadmaps with tables, multi-sheet XLSX part lists, CSV ticket databases, production logs, Python source, photographed labels and diagrams), it answers each question with the exact value and the exact set of source files, or refuses when the corpus does not contain the answer. Architecture: a resident model server starts with the container, loads Qwen2.5-7B-Instruct and Qwen2-VL-2B-Instruct on the AMD GPU (ROCm) and listens on a unix socket, so each per-question process is a thin client well inside the 30-second budget. The index pass walks the corpus defensively (empty directories, unreadable, encrypted and unknown files are logged and skipped, never fatal), parses every supported type, transcribes images with the vision model and persists chunks to disk. Retrieval is BM25 with identifier boosting (ticket ids, error codes, part numbers) plus a second hop that follows rare identifiers across files, which is what makes two-file answers possible. The LLM returns the value and only the passage numbers it strictly needed; citations come from those passages, never from the whole retrieval. Withdrawn or superseded documents are down-weighted and the prompt forbids answering from memory. Why it matters: the author is a Brazilian accountant and court-appointed expert who spends hours locating the right legal norm, service code and tax rate across hundreds of documents. The same engine powers FiscalRAG, a public Brazilian tax-law assistant over 1,059 legal and municipal records, and is the core of Vyson Fiscal Cloud, a compliance SaaS with eight agents: legislation watcher, tax sheet builder, calculators, tax planner, NFS-e issuer, obligations calendar and auditor for tax recovery. Everything is open source (MIT).

Six Matches. One leaderboard. Build on AMD, earn points and compete for prizes.

🌎 Hybrid Hackathon | 💻 Online Build: 3–8 November 2026 | On-site Phase: 8 November 2026 📍 On-site Venue: Pierpont Place 163 W Pierpont Ave, Salt Lake City, UT 84101, USA ✨ Please note that travel and accommodation are not covered.

🌎 Hybrid Hackathon | 💻 Online Build: 12-17 October 2026 | On-site phase: 17-18 October 2026 To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. The AMD Developer Hackathon: ACT III is a hands-on event for developers, founders, engineers and builders who want to push what’s possible with AI on real infrastructure. 📍 If approved for the on-site experience, you'll also have the opportunity to showcase your project live. Travel and accommodation will not be covered.