Farzad Bagheri@FarzadSlash
2
Events attended
2
Submissions made
France
3+ years of experience
About me
Passionate full-stack developer specializing in React, .NET, secure APIs, and AI integrations (GPT, OpenAI). I develop modern, elegant, high-performance web applications focused on user experience—from frontend to backend. My skills include: • Modern frontend development (React, Angular, TypeScript, Tailwind) • Robust backend development (ASP.NET Core, Entity Framework, JWT, etc.) • AI integration (chatbots, intelligent assistants, GPT APIs, etc.) • SQL databases, dashboards, and secure systems
🤓 Latest Submissions
AgentWatch — AI Agent Security Platform
AgentWatch is a real-time AI agent security platform built on Veea Lobster Trap, securing both Google Gemini and Anthropic Claude backends equally. THE PROBLEM: AI agents read files, hit APIs, and trigger actions in production — but guardrails haven't kept up. A single malicious prompt can leak credentials or exfiltrate data. Most teams have zero visibility into what their agents are doing. WHAT WE BUILT: Deep Prompt Inspection: Every prompt analyzed for injection attempts, jailbreak signatures, PII, credential leakage, and exfiltration patterns via Lobster Trap DPI. Risk scores 0.0–1.0 in real-time. Multi-LLM Security: AgentWatch secures both Gemini 2.0 Flash, Gemini 2.0 Pro, and Claude Sonnet. The same Lobster Trap security layer applies to all backends equally — switch models without losing governance. Policy Enforcement: First-match-wins YAML rules compatible with Lobster Trap. Actions: ALLOW, DENY, LOG, HUMAN_REVIEW, QUARANTINE, RATE_LIMIT. HIPAA, SOC2, and finance packs included. Real-Time Dashboard: Blazor WebAssembly plus SignalR — every attack and alert appears live. Risk trend charts, threat breakdowns, live event feed. Full Audit Trail: Every interaction logged with declared vs detected intent, risk score, and threat flags. CSV export for regulators. LIVE DEMO RESULTS: - Prompt Injection: 0.95 risk → BLOCKED - Data Exfiltration: 0.98 risk → BLOCKED - PII Detected: 0.55 risk → LOGGED - Intent Mismatch: 0.78 risk → HUMAN_REVIEW - Clean Request: 0.02 risk → ALLOWED STACK: ASP.NET Core 10, Blazor WebAssembly, SignalR, SQLite, Google Gemini API, Anthropic Claude, Veea Lobster Trap. Lobster Trap is the floor. AgentWatch is the ceiling.
19 May 2026
ContentDNA — Creative Fingerprint Engine
ContentDNA is a multimodal AI application that generates a unique creative fingerprint for any media content. Just like DNA reveals a person's identity, ContentDNA reveals the hidden creative identity of your work. Upload an image, video, or audio file and instantly receive a deep structured analysis: a DNA Score (0–100 distinctiveness rating), a 5-dimensional mood profile (energetic, optimistic, cinematic, melancholic, intense), pacing and rhythm analysis (cut rate, BPM, visual density), a style fingerprint with 8 descriptive tags, color palette breakdown, the top 3 real-world creative references it most closely resembles with match scores, target audience prediction (age range, lifestyle, best platform), and a plain-language AI creative brief. The application is built on a React 18 + Vite frontend with a Node.js + Express backend. Media preprocessing uses ffmpeg for video frame extraction and Sharp for image optimization. The AI vision layer uses Gemini 1.5 Flash for multimodal understanding across image, video, and audio inputs — architected to run on AMD Instinct MI300X via ROCm and Qwen-VL when AMD Developer Cloud credits are provisioned. ContentDNA targets filmmakers, content creators, brand teams, and audio producers who need to understand the emotional and stylistic DNA of their creative work in seconds. The project is fully open-source and deployed as a Hugging Face Space.
10 May 2026
