Flow time: 5 min I your weekly pulse on AI news, tool and case studies reshaping the water sector
🔍 What’s in today’s flow
💧 IWA shows how GenAI and agentic AI can automate reporting, preserve workforce knowledge, and support decisions as utilities face climate and asset pressures.
🌾 Kilimo proves AI-guided irrigation can deliver real water savings, validating over 220,000 hectares across seven countries for corporate water programs.
📌 Plaud launched NotePin S, a wearable AI recorder that transcribes and summarises meetings without bots or screens.
🔍 Syncly uses AI to turn customer feedback into early warnings, helping organisations fix issues before customers leave.
⚖️ Alaska courts found AI chatbots risky for legal advice, after a probate assistant produced confident but incorrect information.
🔬AI research spotlight: GenAI and Agentic AI for water utilities
The details
IWA published a roadmap showing utilities how Generative AI can draft reports and explain complex data in plain English, while Agentic AI runs scenarios, adjusts operations autonomously, and helps executives plan for climate chaos.
Key points
GenAI automates time-consuming tasks including compliance reporting, work order drafting, and natural language explanations of complex operational data
Agentic AI enables continuous scenario testing, proactive risk monitoring, and autonomous adjustments to pump schedules and treatment processes
Five core value areas deliver measurable impact: productivity gains, workforce knowledge transfer, complex decision support, resource efficiency, and executive foresight
Why it matters
This research provides water utilities with a practical roadmap for adopting AI responsibly during unprecedented workforce transition, climate uncertainty, and infrastructure pressure, showing how technology can augment human expertise while improving service reliability and sustainability.
🤖Latest in AI: Meeting notes that don't suck
Source: au.plau.ai
Plaud just launched two products that solve the meeting documentation problem everyone pretends isn't a problem. The NotePin S ($179) is a wearable recorder with 20 hours of battery that clips to your shirt, hangs on a necklace, or sticks magnetically anywhere. The desktop app lurks in the background, detects when you join Zoom or Teams, captures everything without joining as a bot, and turns conversations into searchable, structured notes with AI summaries.
Why it matters
Water professionals spend absurd amounts of time in technical meetings about asset conditions, compliance requirements, and incident responses where missing details create actual problems. This tech captures everything so your engineers can focus on solving problems instead of frantically typing notes. When experienced staff retire, you'll have searchable records of every decision they made and why.
🔧 Case study: Proof that agricultural AI actually saves water
What happened
Kilimo deploys AI irrigation tools across seven countries, telling farmers exactly when and how much to water using satellite data and machine learning. The result: 8 million cubic metres verified saved across 220,000 hectares. Tech giants including Google, Microsoft, and Intel are paying farmers directly for these measurable water savings through transparent, auditable programs that include irrigation conversions and regenerative practices.
Why it matters
Agriculture uses 70% of the world's freshwater, and most irrigation efficiency claims are marketing fiction. Kilimo proves AI can generate real, verifiable water savings that benefit both farmers' wallets and downstream water users. When your utility competes with farms for the same water, projects like this aren't just good news, they're survival strategy.
🔧Trending tool: The customer complaint detector you need
Source: Fondo.com
Syncly is the Y Combinator-backed AI platform that customer service teams are quietly obsessed with. It inhales feedback from support tickets, reviews, social media, and even TikTok videos, then uses AI to spot patterns that predict major problems before customers start canceling or regulators start calling. It's particularly good at catching the signals traditional text analysis completely misses.
Key features
AI-powered auto-tagging instantly categorises incoming feedback across all channels and reveals patterns in customer sentiment
Real-time anomaly detection monitors sentiment trends and flags critical unresolved issues at aggregate and individual account levels
Bi-directional integration with Intercom, Zendesk, Salesforce, and Front enables sentiment data to flow back into workflows
⚖️ AI Tool Scorecard
Ease of use: ⭐⭐⭐⭐ Intuitive dashboards and automated categorisation, though initial setup requires defining custom categories and connecting data sources.
Cost: Enterprise SaaS pricing with annual contracts suits mid-sized to large utilities but may challenge smaller operations.⭐⭐⭐
Security & privacy: Maintains SOC 2 Type 2, HIPAA, and GDPR compliance standards appropriate for customer data.⭐⭐⭐ ⭐
Integration: Strong integration library covers major CRM platforms, though specialised utility systems may require custom API development.⭐⭐⭐⭐
Overall:15 /20 - If you're drowning in customer complaints about water quality, billing errors, or service interruptions, Syncly finds the patterns that tell you what's actually broken in your operation before it becomes a crisis
🕵️AI’s shadows: The $2 million chatbot that couldn't stop inventing facts
Source: nbcnews.com
Alaska courts spent over a year and substantial budget building AVA, an AI chatbot designed to guide people through probate procedures. Despite being restricted to official court documents, the system repeatedly hallucinated, inventing non-existent law schools and giving procedurally wrong advice. After 15 months of trying to fix it, they admitted defeat and pulled it down.
Why it matters
This is what happens when AI confidently delivers wrong answers in situations where wrong answers have consequences. Water utilities face identical risks with AI systems handling water quality guidance, treatment controls, or regulatory compliance. One hallucinated boil water advisory or incorrect chemical dosing recommendation and you're looking at public health consequences that make Alaska's court problems look trivial.
Takeaway
AI adoption requires rigorous testing and human oversight, especially where errors carry real consequences for public safety or legal compliance
Thanks for reading! I hope you’ve enjoyed this week’s edition and look forward to seeing you next week!
Dr. Andrea G.T