

The $3B global freelance market is broken by a massive "Trust Gap." Buyers fear paying for bad work, and sellers fear non-payment. Traditional platforms solve this with slow manual arbitration and rigid 14-day hold periods, stifling economic velocity. ArcWork is the solution: The Economic OS for Agentic Freelancing. We are the first platform where capital, humans, and machine agents coordinate with sub-second finality. By combining Circle's Programmable Wallets with Gemini's multimodal reasoning, we have replaced the "gig economy" friction with programmable, autonomous trust. How It Works: Smart Escrow: The buyer deposits funds into a smart contract on Arc L1 (using native USDC). Multimodal Work Submission: The freelancer submits their work (code, designs, files). Agentic Verification: Our Gemini-powered AI agent analyzes the submission against the original Scope of Work, inspecting text, images, and file contents. Trust-Minimized Release: Success: If the work matches the requirements, the AI triggers the smart contract to instantly release the funds. Risk: If anomalies are found, a "Tactical HUD" alerts the buyer to manually review or request revisions. The Tech Stack: Settlement: Arc L1 for deterministic, low-cost finality. Treasury: Circle Programmable Wallets for secure, developer-controlled fund management. Efficiency: Circle x402 Protocol for "Gasless Validation Credits," allowing freelancers to pre-check their work using AI, off-chain before final submission. Intelligence: Google Gemini 2.5 Flash for high-speed, multimodal semantic understanding of complex deliverables. ArcWork is not just a marketplace; it is a protocol for high-velocity, usage-based services where trust is code.
24 Jan 2026

The Open-Source Multi-Modal Medical AI Assistant is a privacy-first digital health platform that leverages state-of-the-art technologies such as the Groq API, LLaMA-based large language models, Whisper speech-to-text, and computer vision models like EfficientNet and ResNet. It allows users to interact through text, voice, or images to receive preliminary medical information, triage guidance, and skin condition screening. Trained on medical datasets the assistant is designed to be accessible, accurate, and transparent—empowering individuals with reliable health insights while maintaining strict privacy standards. This project aims to bridge gaps in healthcare access, especially in underserved communities.
8 Jul 2025

Our project is an AI-driven mini learning path tool designed to guide users through a personalized educational journey. By allowing users to input a topic of interest, the tool generates a structured, step-by-step learning plan tailored to their goals, current knowledge level, and pace. It curates the most relevant resources—videos, articles, courses—and organizes them into a coherent path, ensuring that users stay on track and make consistent progress. Whether you're a beginner diving into a new subject or someone looking to deepen your expertise, our platform adapts to your learning style and helps you grow efficiently and confidently.
1 May 2025