Most feedback never gets filed as a ticket, it shows up as friction in logs and confusion in conversations. Here’s how we turned scattered signals from errors, usage logs, and direct feedback into a daily digest we actually trust.
Closing the Feedback Loop: From Errors and Usage Signals to Daily Action Read Post »
AI is changing low-code. Logic is moving from visual workflows into generated functions, while diagrams become a way to explain and validate the result.
The Evolution of Low-Code in the Age of AI Read Post »
Most data tells you what happened — but rarely why. At Maranics, we connect machine data with human context, turning incomplete streams into actionable insights. By capturing input through flexible digital checklists and workflows, we make data meaningful, reliable, and AI-ready.
Maranics – The Missing Link for Contextualizing Data Read Post »
We’re evaluating voice input to make maritime workflows more accessible – especially for crews working in constrained conditions. Instead of tapping through checklists, what if you could just talk?
The Future of Hands-Free Operations Read Post »
Discover how reranking vector search results helps our AI agents and services deliver more accurate and relevant information. By reordering search results based on user intent, we improve decision-making, reduce unnecessary data, and ensure our customers always receive the best possible answers.
Reranking Vector Search Results: Improving Accuracy in AI Workflows Read Post »
We benchmarked Jetson Orin Nano, Coral TPU, and CPUs to evaluate real-world AI performance for vessel use. Jetson offered strong results but required tuning; Coral was efficient but limited; CPUs were flexible but underpowered.
AI on the Edge: Evaluating Options for Vessel Applications Read Post »
When we first introduced our AI assistant for workflow automation, we weren’t sure how it would land. But the feedback from
Building Smarter Together: Updates from the AI Assistant and What’s Next Read Post »
