LoopStudio

It's Friday... the week is winding down, and you might be feeling that typical burnout from searching for the right professional fit. If you haven’t found that ideal workspace yet, take this as a sign ‼️ We’re looking for 5 Senior Full-stack Java + Angular Engineers to join us at LoopStudio. You’ll be working with a major partner in Fintech and Retail. This is a space where you can actually lead the modernization of a product and have a say across the entire stack. We’re building a team that values quality and secure code, and we’d love to have you on board. If you're ready for a change that actually feels right, let’s talk!

You can’t fight modern AI risks with legacy tools. Our work with Neon Cyber is a great example of how to apply AI responsibly to protect companies directly from the browser. Neon Cyber uses an AI-powered agent to analyze navigation in real-time, blocking phishing and identifying "Shadow AI" - those unauthorized tools that expand a company’s attack surface. ⬇️ How did LoopStudio help? -We built the data infrastructure to process these AI alerts without latency. -We implemented dedicated AI processing services to categorize unsafe web activities. -We translated complex raw data into a visual dashboard that any manager can understand, significantly reducing cognitive load. For us, AI integration it’s about making the invisible visible so companies can operate securely. This is the future of secure development. AI-driven teams, ready to build with you 🚀

💡 In our latest Tech Council, Federico Iman and Francisco Donadio shared how we integrate AI tools like Cursor into our daily workflow at LoopStudio. 🧑💻 Key takeaways from the session: -AI isn't a silver bullet for legacy systems: Resolving bugs in older projects requires more than a general prompt. Federico demonstrated that the developer must provide specific context - failed endpoints and API responses - to ensure the AI follows project conventions without unnecessary refactoring. -Guardrail the business logic: To prevent AI from breaking critical functions, Francisco uses "Skills". These are documented "golden rules" (time calculations, rounding, integrations) that the AI must consult before proposing changes. -Purposeful automation: We are connecting our workspaces to Notion via MCPs to turn tickets into technical plans automatically, saving manual setup time while keeping the human in the loop. 👉 The bottom line: AI gives us incredible speed to squash bugs, but it doesn't always have the full picture. Real value comes when the team sets the rules to protect what already works. What do you think? ⬇️

We’re looking for new talent to keep building products that combine security, design, and technology 🚀 Apply today or share this post with your network.

Renato Cargnelutti, one of our Tech Leads, recently led a session on the transition from DevOps to DevSecOps. 💡 The core message was clear: security isn't a "final gate," but a continuous process that starts the moment we design the architecture. What this means for our team 👇 -Shift Left: we integrate security early - from threat modeling to automated testing - to catch vulnerabilities long before production. -Shared Responsibility: security is everyone's job, not just a specialized team. -Beyond Silos: by sharing knowledge on tools like Snyk and SonarQube, we ensure our entire team speaks the same technical language. At LoopStudio, we don't just build software; we invest in the collective expertise that makes it secure 🤝

Over the past year at LoopStudio , we’ve been exploring AI agents and building practical use cases using tools like LangChain, LangGraph, and MCP. Recently, I took this DeepLearning.AI course by Andrew Ng to revisit some of the core concepts and design patterns behind agentic AI. It’s framework-agnostic, hands-on, and does a good job of clearly explaining ideas that are easy to blur together when you’re just building. Some of the topics covered: • Evals: How to evaluate your agent in both objective and subjective ways, including LLM-as-a-judge approaches. • Reflection design pattern: Improving task outputs by using additional agents to inspect inputs and outputs. • Tool usage: Syntax, function calling, and how to structure and expose your own functions as tools. • Practical trade-offs: When it makes sense to optimize for cost and latency plus how to measure workflows to identify bottlenecks. • Planning pattern: Creating plans and executing them step by step. • Multi-agent patterns: Coordinating execution across multiple agents: sequentially, many-to-many, or hierarchically. If you're working with agents, this is a useful way to step back and understand the underlying patterns a bit better.

🧑💻 Meet Marcos Martinez, our CTO! We realized we talk a lot about code and security, but we rarely show the person making the big technical calls at LoopStudio. Marcos explains how we actually help our partners, from the first Design Sprint to scaling a product. It’s a good summary of what we do, and it reflects all our teams working to make our architecture and security more robust every day 💪 If you’re curious about how we’d approach your specific project, feel free to drop a comment or message Marcos directly.