Trend Radar

Inspiration

Every creator we know spends hours each week scrolling Reddit, Hacker News, and Twitter trying to figure out what to make next. Half the time they end up covering something they've already talked about. We wanted an agent that does that research for you — automatically — and tells you exactly where the gaps are.

What it does

Trend Radar is an autonomous content intelligence agent. It scans live trending topics from Reddit, Hacker News, and GitHub every 60 seconds, cross-references each trend against your content history stored in Senso.ai, and flags uncovered opportunities with AI-generated content briefs. You can click any trend to open a strategy chat that helps you plan angles, outlines, and formats. You can also ingest your past content so the gap detection gets smarter over time.

How we built it

• Railtracks for agent orchestration — manages the multi-step pipeline of fetching trends, checking coverage, and generating briefs
• Senso.ai as the knowledge base — stores the creator's content history and enables semantic search for coverage detection
• assistant-ui for the chat interface — React primitives for threads, composers, and message rendering
• Unkey for API key verification on every route
• Gemini 2.5 Flash as the LLM powering the strategy chat and brief generation
• Next.js 16 with a Python FastAPI backend, deployed on Vercel

Challenges we ran into

Getting the Railtracks agent to return clean, parseable JSON from the LLM was tricky — we had to build a response cleaning layer to strip wrapper objects. We also hit hydration mismatches in React when adding the auto-scan countdown timer, since the server and client rendered different initial states. Swapping from Anthropic to Gemini mid-hackathon required rewiring both the frontend and the Python agent.

Accomplishments we're proud of

The auto-scan loop — Trend Radar runs fully autonomously without any manual intervention. You open it and it just works, rescanning every 60 seconds and updating the dashboard in real time. The coverage detection is also surprisingly accurate even with a small knowledge base.

What we learned

How to wire together multiple sponsor tools into a cohesive agent pipeline. Each tool handles a specific piece — Railtracks orchestrates, Senso stores knowledge, Unkey secures the API, assistant-ui renders the chat — and the result is greater than the sum of its parts. We also learned that the best hackathon demos solve a problem the judges can immediately relate to.

What's next for Trend Radar

• Social media ingestion — paste your TikTok, YouTube, or LinkedIn URL and it auto-imports your entire content history
• YouTube Trending and Twitter/X as additional trend sources
• Scheduled scans with email/Slack notifications when a high-score gap appears
• Publish pipeline — push the generated brief directly to Notion, Google Docs, or your content calendar
• Niche filtering — configure your content niche so it only surfaces relevant trends

Built With

• gemini
• railtrack
• senso.ai