The transition from monolithic systems to modular, edge-based architectures represents one of the most significant transformations in modern software development. The BLT project, initially built on Django, has undertaken a strategic migration to Cloudflare Workers, breaking down its ecosystem into independent, fast, and lightweight products. Each of these modules — from the design system and documentation to the newsletter manager and hackathon panel — now has its own frontend, API Worker, D1 database, and authentication flow. This rethinking not only improves performance and ease of deployment but also democratizes contributions by making each piece understandable and autonomous. In the business realm, this philosophy aligns with the best practices of custom applications that we promote at Q2BSTUDIO, where we design custom software that can scale frictionlessly thanks to decoupled architectures.
The BLT migration illustrates how abandoning a heavy backend allows unleashing the potential of each functionality. For example, the authentication module has been separated into two layers: one based on API keys for integrations and another with JWT for user sessions, facilitating the incorporation of AI for businesses and AI agents that require secure identification. Visual consistency has become a fundamental pillar; instead of each repository inventing its own interface, a shared design system has been established with reusable patterns (colors, cards, layouts, dark mode). This is reminiscent of how at Q2BSTUDIO we apply cloud services aws and azure to ensure applications maintain a unified experience even when distributed across multiple microservices. Additionally, centralized documentation and repository maps lower the barrier to entry for new developers, a critical aspect when working with distributed teams and cybersecurity as a cross-cutting requirement.
One of the most emblematic cases is the newsletter module, which evolved from a simple static flow based on GitHub issues to a full platform with subscribers backed by D1, admin authentication, send queues with SendGrid, and dashboards via API. This type of transformation parallels the business intelligence services projects we implement, where consolidating data from multiple sources — such as power bi for visualization — requires a lightweight yet robust infrastructure. Similarly, the timer module and the daily check-in module (Sizzle) have been redesigned to eliminate external OAuth dependencies, adopting their own authentication and Python backends on Workers, which reinforces independence and security. These patterns are directly applicable to the development of custom applications that we offer at Q2BSTUDIO, where each functionality is built as an independent yet integrable product.
The second half of the BLT project will focus on hardening and completing this ecosystem: polishing authentication, removing legacy assumptions from Django, improving deployment guides, and finalizing the CLI that unifies the developer experience. This modular foundation approach is exactly what we recommend to our clients when they seek artificial intelligence integrated into their business processes or AI agents operating on serverless infrastructures. The main lesson from this migration is that changing the runtime forces a complete rethinking of the architecture, but the benefits — speed of delivery, ease of maintenance, and a more active community — far outweigh the initial effort. At Q2BSTUDIO, we apply these principles day in and day out to transform legacy systems into modern, scalable, and secure platforms, whether through cloud services aws and azure or through the development of custom software that adapts to the changing needs of the market.
