Sauron

Inspiration

We built Sauron around a simple insight from the TruckerPath challenge: the driver side of trucking has increasingly strong digital tools, but dispatch is still stuck operating in an intelligence gap. Small and mid-sized fleets often run on phone calls, spreadsheets, fragmented updates, and tribal knowledge. That means load assignment, safety monitoring, and cost control are often reactive instead of proactive. We wanted to build the AI layer that gives dispatchers true real-time situational awareness.

What it does

Sauron is an AI-native fleet digital twin and vision copilot for trucking operations. It gives dispatchers a real-time control tower with live fleet visibility, AI dispatch recommendations, assignment simulation, cost insights, streaming fleet chat, proactive alerts, and SAURON Vision for camera-based safety monitoring. Instead of reacting after something goes wrong, dispatchers can see which driver is best for a load, understand cost impact before assigning it, monitor route and operational risk, and catch issues like fatigue, distraction, or road hazards earlier.

How we built it

We built the frontend in Next.js, React, TypeScript, Tailwind CSS, Zustand, Mapbox GL, and Recharts to create a fast, real-time dispatcher experience. We used InsForge, PostgreSQL, and Edge Functions to manage fleet data and power the core AI workflows for dispatch recommendations, cost intelligence, simulation, alerts, and SAURON Vision. We also used a local FastAPI service for localhost operational workflows and historical queries. Claude Sonnet powers the AI across chat, dispatch reasoning, and vision analysis. A big part of the build was making these features feel like one coherent operating system rather than separate demo modules.

Challenges we ran into

One major challenge was making the product cohesive. Live mapping, AI chat, dispatch scoring, simulation, alerting, analytics, and camera-based safety monitoring are all useful on their own, but the real challenge was tying them into one dispatcher workflow. Another challenge was reliability: vision analysis is powerful, but hackathon demos need resilience, so we built deterministic fallback logic to keep SAURON Vision operational even when AI analysis fails. We also had to coordinate a hybrid architecture across frontend, backend services, and serverless functions while keeping the app responsive and polished.

Accomplishments that we're proud of

We’re proud that Sauron addresses four of the challenge’s most important problem areas in one experience: Smart Dispatch, Proactive Alerts, Cost Intelligence, and Safety & Compliance. We’re also proud that it feels like a real product rather than a hackathon prototype stitched together from separate parts. The map, dispatcher workspace, simulation flow, reporting layer, alerting, and vision monitoring all work together to tell one clear story: helping fleets act earlier and operate smarter. Most importantly, Sauron fits naturally into the TruckerPath vision by strengthening the connection between drivers in the field and dispatchers making operational decisions.

What we learned

We learned that AI in logistics is most useful when it improves operational decision-making, not just when it generates answers. Dispatchers need clarity, speed, and trust. We also learned that combining real-time data, predictive workflows, and AI vision requires strong fallback systems, clean state management, and a product design that keeps every feature anchored to a real operational decision.

What's next for Sauron

Next, we want to connect Sauron to richer live fleet telemetry, expand alerting and intervention workflows, and deepen billing and document automation so it supports even more of the dispatch-to-cash lifecycle. We also want to evolve SAURON Vision from demo monitoring into a more robust real-time fleet safety layer. Long term, we see Sauron becoming the intelligence layer on top of TruckerPath’s network: not just showing what is happening, but helping fleets decide what should happen next.