fiberous

@kyle_e_walker you are making it very hard to justify spending on a BA license

Another update to R's spopt-r package: you can specify "required" and "candidate" facilities in max coverage / location-allocation solvers Brief demo: choosing the best new fire stations from candidate points in Fort Worth Same functionality here as ArcGIS's Network Analyst

@seanpatward Surprised green machine lets yall do that

Holy smokes, so many people still don’t get this. Let’s say a client is going to hire someone to lease or sell their property. They are going to pay the same market fee. Do they choose: - a solo broker (who may or may not have an assistant)? - a solo broker who is leveraging AI to be 5x more productive? - a broker who leverages human capital and has a team of ten people? - that same team of ten where each person is leveraging AI to be 5x more productive? This seems so obvious to me. For the same cost I’m picking the team that brings 10-50x the horse power. Our team is all-in on leveraging AI to the benefit of our clients.

@JakehellerAI @jshirsh1 FWIW I’ve gotten much better mileage pointing Claude at the API docs and having it write a SKILLS doc mapped to endpoints. Doesn’t miss as much and keeps context window clean

@jshirsh1 Custom mcp server

Here are all the tools I've connected to claude cowork so far: 1) Gmail 2) Google Calendar 3) Pipedrive (my CRM) 4) Granola (AI notetaker) 5) Notion 6) Gamma 7) Zapier asked it how's my pipeline looking? 35 deals. Health score: 22/100 (ouch) then asked for action items from that days calls then set up a scheduled task to scan my email every morning for new deals and underwrite them Pretty fired up on claude connectors...

@JJWatt If staff doesn’t bring me water when seated, 0%

Genuine question on a restaurant situation: You walk up to a counter to order. You find your own table and seat yourself. If you order coffee, they hand you a cup and you go fill it up yourself. If you order food, they hand you a buzzer and when it goes off, you go pick it up yourself. The iPad has a “20%, 25%, 30%, Other” tip option, with 20% already preselected. What’s your move?

@toddsaunders Never heard of a reverse join. Is that essentially building a training set to score universe against?

I built a land acquisition intelligence/sourcing platform that scores 1.5M parcels of land across the I-95 corridor for data center and industrial conversion potential. All with Claude Code... in less than 36 hours. This used to take their team weeks. Here's how I did it. 1/ Ingested 1.5M parcels from NC OneMap's ArcGIS REST API with full polygon geometry, zoning, ownership, and tax records across 14 counties 2/ Pulled I-95 corridor geometry and 142 interchanges from OpenStreetMap's Overpass API 3/ Ingested 1,107 substations, 8,174 transmission lines, and 292 gas pipelines from HIFLD federal infrastructure data 4/ Ran PostGIS distance calculations on every parcel to I-85 centerline and nearest interchange using LATERAL KNN joins (100x faster than cross joins) 5/ Normalized hundreds of raw county zoning codes into six categories and scored farmland confidence 0-100 using tax programs, land use codes, and acreage heuristics 6/ Built a motivated seller detection engine: out-of-state owners, estates/trusts, tax delinquency, long hold periods, declining assessed values 7/ Calculated conversion readiness scores using reverse-join spatial queries against 20K industrial parcels (turned a multi-hour query into 15 seconds) 8/ Scraped 2,100 active listings from Redfin's public CSV endpoint with recursive bounding box subdivision to beat their 350-result cap 9/ Wired it all into a composite acquisition score (0-100) with configurable weights, a full-screen map explorer with vector tiles, and a document generation suite that produces institutional-grade investment memos, slide decks, and automated intelligence briefs The whole thing runs on PostGIS, FastAPI, React, and MapLibre GL JS. I did all of this with no paid mapping APIs and no paid data sources... at least not yet. Every piece of spatial math runs inside the database, not Python. That's how you process 1.5M parcels without melting your laptop. And you still want to tell me bespoke software isn't the future?

See below for how cooked you are/will be

@JakehellerAI Makes sense, forgot you were a dev. So in a way you are seeding the demise of everyone you work with downstream without being affected by consequences. Not a moral judgement just an observation.

@xplosivememory I'd be a lot more concerned about replacement if I was an employee

Here are claude cowork possibilities for CRE as of today (whether they are practical for you, you decide): 1) Scan your inbox every morning for new OMs and auto underwrite them 2) Query your entire deal pipeline with plain english ("what deals need my attention today") 3) Pull action items from every meeting you had this week in seconds 4) Get a daily briefing that combines your calendar, email, CRM, and meeting notes into one summary 5) Score your pipeline and flag stale deals, missing close dates, and zero activity opportunities 6) Generate decks (investor decks, BOVs, OMs) using Gamma without leaving Claude 7) Deep dive your meeting notes for insights you missed and surface new ideas 8) Draft personalized outreach emails based on prospect research across LinkedIn, web, and your CRM 9) Create and update tasks in project management software (for me I use notion) 10) Check your Google Calendar and find meeting conflicts or open slots 11) Auto create follow up tasks after every call based on your Granola (or other AI notetaker) notes 12) Search your email history for how you underwrote a similar deal in the past 13) Search your drive to find most widely used contracts (PSAs, LOIs...etc) 14) Pull comps from software providers and create reports + presentations (all within claude) 15) Build a weekly pipeline report and email it to your partners every Monday 16) Connect DocuSign and track which LOIs/PSAs are still unsigned 17) Query Yardi for data inside claude (yardi virtuoso MCP connectors are in active rollout!!!) 18) Create Canva marketing flyers from deal data without switching apps 19) Summarize every call from the last week and rank them by importance (assuming you used an AI notetaker ) 20) Set up a scheduled task that checks for lease expirations across your portfolio each week 21) Set up a scheduled task to check for new permits pulled 22) Pull your Stripe or payment data to track which clients are current on fees 23) Send iMessages or Slack updates to your team when a deal moves stages in your CRM 24) Build a morning routine that checks email, calendar, CRM, and news in one shot 25) Query your Google Drive for every document related to a specific property address 26) Create custom sales plugins so your whole team can run the same pipeline commands 27) Use browser automation to enter deals into your CRM 28) When you get sent a full due diligence room (T12, rent roll, leases, disclosure...etc), underwrite the deal, create DD checklist, identify red flags and back into purchase price based on your criteria 29) Scan your drive/emails to surface the PSA you most commonly used to transact with (to help mitigate attorney costs) 30) Every morning, pull maintenance requests from your email (or management software), categorize by urgency 31) Pull every new lease and automatically create lease abstract in excel 32) Pull deliquency data from Yardi (also appfolio I believe through custom MCP server), draft past due notices, and schedule follow up reminders 33) Pull vendor invoices from email, compare against budgeted amounts, flag anything over 10% 34) Every week pull your draw schedule from your spreadsheet and compare against actual invoices received through email 35) Search your emails for every LOI or counter offer you sent + underwriting (in a specific market), create your criteria, and apply it to future deals The future is bright! Sometimes I have to pinch myself as a reminder that we're looking at the worst version of these tools today. It will get better, faster, and easier to use... The use cases are endless... Just be thoughtful about spending time on practical applications. Some of these (among many others) may not be the best use of your time.

@kyle_e_walker What package gives you this isochrones, looks clean & snappy as hell

With OpenClaw, you can have on-demand location intelligence in your pocket. I taught my bot how to pull demographic data and do drive-time analytics for any address using validated pipelines; It then formats a branded PDF report and delivers to me by Telegram, ready to share.

@JoaquinCastrotx All victims obeyed the posted sign no guns allowed sign, how did that work out for them

A horrific shooting in Austin early this morning outside a popular bar took the lives of three people and injured fourteen more. No one should fear facing violence on a night out. The victims and their families are in my thoughts and I am praying for the recoveries of those hospitalized. Congress must continue to work to end the scourge of gun violence in our country.

@SarahEckhardtTX Wow are you saying immigrants shouldn’t be afforded 2nd amendment rights?

Austin, known for its vibrant and peaceful nightlife, experienced deadly violence again last night. My thanks go to those who intervened, and my heart goes out to those injured and killed. The gunman is believed to have had a pistol and an assault rifle. We must address the consequences of having made deadly weapons so common and accessible.

@kyle_e_walker Did you design the pipeline direct in Claw or in Opus then port it over

Checking in with my OpenClaw bot about the latest Texas drilling permits, still working great. Yes, you can do this with a cron job (my bot executes an R/Python job here). What's useful is the interactive piece - I can ask follow-up questions, have the bot write a summary, etc.

@kyle_e_walker Context window and OCR are unmatched

For me, Gemini is the #1 LLM for several use-cases including document extraction and image generation. But it doesn't get nearly the hype of its competitors as the Gemini CLI is a distant third for agentic coding. For us users - it pays off to know the best tool for the job.

@kyle_e_walker I feel vindicated for never bothering with arcpy or model builder

@xplosivememory if you "teach" the LLM a well-built framework first (ArcGIS JS SDK, R's mapgl, raw deck.gl, etc.), you can do absolutely incredible stuff these days

Working on visualizing the ERCOT interconnection queue. 1,880 active projects, ~415 GW in the pipeline - roughly 5x Texas peak demand. Fuzzy-matched queue entries to substations from their points-of-interconnection with fallback to county when necessary.

@TristenPalori So why do they need a broker at all?

I just paid $200/month for an AI tool most people don't know exists. After today, I'm never canceling. It turned 5-6 hours of work into 45 minutes. Here's 4 projects I used Perplexity's "Computer" mode for today as a Commercial Real Estate Broker: (#4 is insane)

@HealthcareREguy Ok this one is pretty good:

@xplosivememory Haven’t read this yet x.com/michaelxbloch/…

The most thought provoking article I’ve read in a long time, maybe ever. Really do yourself, your family, and your finances a favor and read it. I don’t think everything in this article will come true, but only a fraction is necessary to turn the world economy on its head.

@HealthcareREguy I’m a little sus of any thought pieces written by tech guys, rose tinted glasses etc Have you read If Anyone Builds It Everyone Dies?