Skip to Content

6 months - What’s New with Daaisy? 🤖 

  • Uploading documents capability: Users can now upload documents directly into the Daaisy interface and keep working 📤⚙️. Daaisy has the power to interpret documents such as ACT leases or title documents. So Daaisy can find, comprehend and contextualise ACT lease purpose clauses, which is crucial for interpreting lease rights in the ACT. ✦🦢
  • Instant Document Intelligence: You can now upload any document within a session and query it immediately. This is a game-changer for workflows like the DA Assist Crown Lease analysis, where you need answers from complex documents in seconds.
  • A Self-Updating Library: The BOK (planning documentation library) is now automated 📚🔄, ensuring the most current legal, policy, and regulatory information is always at your fingertips through engaging with Daaisy. 
  •  Spatial optimisation: Some upgrades have been made to our ‘data core’ to improve basic site facts and geospatial information. This means Daaisy can now understand both text-based information, and details databases, all from one query.


May - What’s New with Daaisy? 🤖

DAAISy is now a reasoning agent! 🧠 http://daaisy.io

We’ve made a fundamental shift in how DAAISy handles your queries:

  • Before: DAAISy used a classifier (fixed routes). Great for patterns, but struggled with complex, mixed queries. 📉
  • Now: DAAISy has an LLM Planner. This reasoning layer dynamically determines which data and documents to retrieve, and in what sequence, based on your specific context. 📈

You can now see the "thinking" in real-time as DAAISy streams her planning steps. No more guessing—just transparent, source-backed planning intelligence (and zero AI slop!). 🚫


April - What's new with Daaisy?

Based on feedback from our professional users group, we have made the following updates to Daaisy:

  • Smarter Location Search (most impactful for users): Daaisy can now reliably find addresses and Section & Block information from natural language descriptions — without requiring users to provide other details which Daaisy should be able to query itself. Previously, approximate address inputs would fail to match. The underlying geospatial query logic has been rebuilt with better fuzzy matching and result ranking.
  • System Prompt Optimisation Daaisy correctly understands its own capabilities — including access to the geospatial database and document knowledge base. It no longer asks users for information it can retrieve itself, and gives more confident, accurate responses about ACT planning matters.
  • Under the Hood Several internal improvements were made to RAG retrieval reliability. These don't change the visible experience directly but improve consistence