正在加载视频...
视频加载失败
I am excited to introduce OpenGeoAgent, a powerful open-source multimodal AI agent for automated geospatial analysis and visualization! It supports QGIS, Jupyter notebook, and Python scripting. In this tutorial, you’ll learn how to automate GIS workflows using natural language, generate maps, analyze satellite data, and even run complex hydrological... show more
13 条评论
This is the perfect example of where the industry is heading. General LLMs are cool, but specialized experts like this are what drive real value. We’re building @uphivexyz to be the home for agents exactly like this
QGIS + Jupyter + Python scripting is exactly the stack that produces brittle, siloed notebooks. An agent traversing all three without handoff failure would genuinely compress months of GIS workflow. The multimodal spatial reasoning is where this actually gets tested.
@SentientAGI Thank you for sharing. I have not tried it out yet. Will look into it to see if it can be integrated into GeoAgent.
Location is only one layer of context, timing and intent often matter just as much🌍🙉
Does this have US streets layers available?
QGIS reminds me of my x DP
QGIS + Jupyter + Python in one agent is the part that stands out. That workflow glue is usually where projects fall apart.
interesting! Question - will this work with open weight LLMs, hosted on premise / hyperlocal for agroforestry, critical city infra such as energy, HVAC, Water, Urban Cooling. Any suggestions what could be that LLM?
Bookmarked
This will come in handy
Cool 👍
Not sure this will help. I am trying to look at global and systemic issues and actually solve them. What does it cost? Is Google Earth Engine already monetized? "open source" is so vague now. What it comes down to "Does this guy make frugal and useful choices so time spent learning his personal selection of tools and conventions is not a huge waste of my time?" It is "introduce" which means he wants money or fame or both. A front end for Google Earth is likely steering people into a honey pit. === How large is the community of Python? and how large is Jupyter community on the Internet? This post was dumped on my time line by X, and it has this glaring bright image that decades of GIS and satellite image data work makes me focus on. But he is another academic coming up with a cute or useful interface that he works on part time hoping it will lead to something. It is based on tools (including python and Jupyter) that have many uncontrollable dependencies and a long long heavy learning curve. It completely hides and diminishes the actual use of these kinds of problems for solving world and country scale systemic problems or opportunities. From my point of view with the Internet Foundation, X acts shamefully not giving me control over my X experience with a "works for me" AI, and then this comes on the poster's term, NOT as I want to see it -- with practical details and costs and global implications. We human X users, need AI that works fr us, on X. And people who push GIS and satellite mage pretty pictures need to learn that is the last thing to show -- if the purpose and use are not clear to ALL potential Internet users -- not just the choir or hopeful job and gig seekers. === 8,300,000,000 humans 6,000,000,000 Internet users 250,000,000 daily X users 23,000,000 Python 47,000,000 Software developers These Jupyter users are hoping it gives them tools to do their work, and that it gives them an edge in jobs and gigs. That is NOT what changes in the world. Do you understand? It is a selfish motive to choose tools, but it is an and servants attitude that changes the world? If one only gathers more and more tools its is boasting and one-ups-manship. If those groups would actually track what they are doing in the world - as an entire community -- then it might have actual systemic value. If they actually see and understand what they are doing as a group -- NOT gamed by a few - it might matter.
awesome not having a job anymore
