.png&w=828&q=75)
(if facing any video playing issues please see it in YT https://youtu.be/JIu-IRZctgY at 4k) This is a retail content pipeline where merchants pay USDC on Arc testnet to unlock AI generation. In Merchant Studio, a merchant uploads product reference images, selects category + variants, then completes payment either via Circle Wallets test payer (developer-controlled) for demos. The backend treats the resulting Arc txHash as payment proof and verifies it via Arc JSON-RPC (receipt success + USDC transfer logs / recipient checks) before starting generation. After verification, we start generating the product images and a looping video using gemini models, persists inputs + outputs to Supabase Storage (stable URLs), writes job logs/receipts/product metadata to Supabase Postgres, and auto-publishes the product to the Storefront for instant browsing + video preview. Circle Product Feedback: Products used: Arc (testnet), USDC, Circle Wallets. Why: Arc+USDC gives a single settlement layer + stable pricing; Circle Wallets enables secure tx execution (challenges) and a repeatable sponsored demo mode. Worked well: Challenge-based transfers; developer-controlled sponsored payments; obtaining txHash for explorer-based proof. Could improve: more guided/typed errors around initialization/PIN + challenges.
24 Jan 2026

(if facing any video playing issues please see it in YT https://youtu.be/7CJ3GqxGJ4Q at 4k) MakeMyCv is an AI-native CV builder that turns a developer’s existing online footprint into a polished, ATS-friendly resume without starting from a blank page. Instead of asking users to manually rewrite their GitHub, LinkedIn, and X profiles into different formats for every job, MakeMyCv treats resume creation as a workflow problem: it intakes, understands, decides, reviews, and delivers in one automated pipeline. How it works: - A developer simply enters their GitHub username, LinkedIn URL, X handle, and target job description into the web app. - The system then fetches public repos, social posts, and profile details, analyzes their tech stack and experience, and generates a structured CV draft tailored to the job. - Behind the scenes, Opus workflow breaks this into focused steps: analyzing repositories, summarizing projects, cleaning LinkedIn experience, mapping skills, and assembling everything into a consistent resume structure. - In the backend Opus Workflow, the CV passes through an AI + human review loop, which checks for ATS compatibility, clarity, and relevance to the job description, while keeping a clear record of which inputs led to which outputs. - Finally, a custom agent converts the reviewed CV into LaTeX, and the backend compiles it into a professional PDF using a single, battle-tested developer resume template. From the user’s perspective, they get a ready-to-use CV plus the LaTeX source, with options to preview, download, and live-edit the document. From a systems perspective, MakeMyCv is a reusable, auditable Intake → Understand → Decide → Review → Deliver automation.
19 Nov 2025
.png&w=828&q=75)
Content Hub is an innovative platform dedicated to transforming social media engagement through the generation of personalized, SEO-optimized content. Our solution is built on advanced artificial intelligence, leveraging IBM’s Granite AI models and their robust APIs for inference. This integration allows us to deliver dynamic digital content that not only captures the unique voice of each user but also meets the stringent demands of today’s competitive digital landscape. By processing data from major social media platforms such as Twitter, LinkedIn, and Instagram, Content Hub converts raw user interactions into compelling narratives within seconds. At its core, Content Hub empowers users with actionable insights and analytics, enabling them to optimize their online presence and refine their digital communication strategies. Our platform exemplifies the potential of strategic AI integration, setting a new benchmark for efficiency and precision in digital marketing. With IBM’s Granite AI technology as our foundation, we ensure reliability and performance while pushing the boundaries of automated content creation. We have meticulously designed system prompts for each mode—Twitter tweet generation, threads generation, LinkedIn post generation, and our analysis models—to maximize output quality.
23 Feb 2025