OmniDoc — Talk to Any Document

Documents aren't just text. Financial reports live in charts. Scientific insights hide in figures. Legal risks bury in tables. Traditional document AI treats visuals as noise. OmniDoc treats them as signal. OmniDoc is a multimodal document intelligence platform that understands everything: text, charts, tables, diagrams, handwritten notes, scanned pages, equations, and mixed-language content. Upload any document and talk to it. Ask: "What was the gross margin trend from section 3 charts?" → OmniDoc reads the bars, not just surrounding text. "Which appendix clauses exceed $500K?" → Parses tables precisely. "Explain the page-12 diagram's relation to the conclusion" → Understands figures in context. Powered by a two-model pipeline optimized for AMD MI300X: • Llama 3.2 Vision 90B processes pages as high-res images, preserving layout and visuals • Qwen3-VL extracts structured data from tables/forms with cross-lingual precision Both run simultaneously on a single MI300X (192GB HBM3, 5.3TB/s bandwidth)—eliminating the complex multi-GPU parallelism H100s would require. Pipeline: 300 DPI page rendering → Llama for semantic structure → Qwen for table precision → retrieval layer → intelligent query routing → cited responses with confidence scores. Performance: 100-page PDF in 42s | 340 pages/min batch | 12 concurrent sessions | ~18× faster than cloud CPU. Use it for: M&A due diligence, regulatory review, academic literature synthesis, contract portfolio analysis, insurance claims with form+image understanding. Ships as a ready-to-use web app: drag-and-drop upload, conversational Q&A, document navigation, and citation tracking that links every answer to its source page and element.

Boundary Forge

Boundary Forge is a model-agnostic AI safety pipeline that helps enterprises deploy LLMs with measurable confidence. Instead of relying on manual red-teaming or hoping a system prompt is enough, Boundary Forge automatically attacks a model, identifies where it behaves unsafely or inconsistently, and converts those discovered failures into runtime guardrails. For this hackathon, we demonstrated Boundary Forge using Qwen 2.5-72B on AMD Developer Cloud with AMD MI300X. Qwen powered the adversarial red-team workflow and was also the model under test, allowing the system to expose real behavioral failure boundaries such as jailbreak attempts, policy drift, unsafe financial guidance, KYC bypass, fraud patterns, coercion signals, asset concealment, and inconsistent refusals. The pipeline works in five stages: generate adversarial probes, run high-throughput model inference, mathematically detect boundary failures, compile those failures into semantic safety rules, and enforce them through middleware before risky prompts reach the LLM. This creates a practical enterprise safety layer that can block, flag, or ask for clarification in real time. The important point is that Boundary Forge is not tied to one model. Qwen 2.5-72B was used to demonstrate the system, but the architecture can benchmark and harden other open-source or proprietary models as well. The goal is to improve models exactly where they fail and make model evaluation repeatable across different deployments. In our AMD Cloud production run with Qwen 2.5-72B, Boundary Forge generated 1,009 unique adversarial probes, fired 4,036 total inferences, discovered 25 boundary failures, and compiled 15 semantic safety rules. The middleware intercepted 68% of known attacks and reduced the effective failure rate from 2.48% to 0.79%. Boundary Forge turns AI safety into an automated engineering workflow: attack, measure, learn, protect, and benchmark again.

Thor v2 — RAG-Free Fitness Intelligence

Thor v2 is a domain-expert fitness AI built on a single fine-tuned Qwen3-8B model trained on 7,118 carefully constructed instruction-response pairs spanning exercise science, nutrition, programming, injury screening, and population-specific guidance. Unlike RAG-based fitness apps that retrieve documents at query time, Thor v2 encodes knowledge directly into model weights during supervised fine-tuning on AMD MI300X hardware using ROCm. Evidence is referenced through compact citation keys — e.g. [CITE:NSCA_HYPERTROPHY_VOLUME] — that the model emits inline. A lightweight citation resolver validates these keys against a locked registry and surfaces the source document on demand. If the model emits an unknown key, it is rejected at runtime. Hallucinated citations are structurally impossible. The dataset covers 113 unique citation keys from 9 authoritative organisations — NSCA, ACSM, ISSN, NASM, HHS, USDA, NIH, CDC, and ExRx — with 80 exercise technique entries and 14 population profiles including senior, postpartum, teen, vegan, rehab return, and competitive athlete. Six conversational style variants (casual, research-nerd, anxious, skeptical, verbose, follow-up-first) are baked into training so the model adapts tone naturally without prompt engineering. Training results: 100% JSON contract pass rate across all eval prompts. Coach gating behavior confirmed — model asks clarifying questions before prescribing when context is missing, rather than giving generic advice. All responses emit valid citation_keys, follow_up_questions, and safety_notes fields. Adapter size: <350MB on top of a frozen 8B base. Built entirely on AMD MI300X (192GB HBM3, ROCm 6.3) using HuggingFace PEFT + TRL. One model. No retrieval. No vector database. The model knows. The resolver proves.

Agronomist

Resume Compact

Resume Compact, the HR Resume Analyzer, is an intelligent, end-to-end recruitment automation platform designed to streamline candidate screening and evaluation by combining modern document processing, artificial intelligence, and web technologies to eliminate manual resume review bottlenecks and provide data-driven hiring insights to recruitment teams. Key technical features include seamless fallback PDF processing with transition from native text extraction to OCR when needed, LangChain structured output using StructuredPrompt and Pydantic to ensure JSON-compliant responses, strict schema enforcement that validates all outputs against Pydantic models and rejects malformed responses, detailed and context-rich prompt engineering to guide the LLM toward accurate structured analysis, async HTTP using httpx for efficient non-blocking calls to the LLM API, robust exception handling with informative HTTP status codes and error messages, and comprehensive audit logging with timestamps, file IDs, and LLM raw responses stored for compliance and debugging. The system reduces recruitment cycle time by 60-80%, eliminates subjective bias in initial screening, allows HR teams to focus on relationship-building rather than document review, and, by standardizing candidate data into structured JSON, enables downstream analytics, reporting, and integration with ATS (Applicant Tracking Systems).

Swarm

Swarm est une architecture multi-agents avancée conçue pour automatiser les workflows techniques critiques, de l'analyse de code à la génération d'infrastructure. En exploitant la puissance et la bande passante mémoire de l'AMD Developer Cloud, cet 'essaim' d'agents synchronisés analyse les dépôts de code pour générer instantanément des configurations conteneurisées (Docker), des scripts de déploiement et une documentation technique complète, accélérant radicalement le cycle de livraison logiciel. En d'autres termes, Swarm transforme des problèmes complexes en tâches modulaires gérées par des agents spécialisés, orchestrés pour atteindre un objectif commun : l'automatisation intelligente du développement logiciel. Imaginez une équipe de développeurs virtuels, chacun expert dans un domaine précis, travaillant en parfaite harmonie grâce à une communication et une coordination optimisées. C'est l'essence de Swarm. Cet outil permet aux équipes de se concentrer sur l'innovation plutôt que sur les tâches manuelles et répétitives, réduisant drastiquement les délais de mise sur le marché et améliorant la qualité du code.