Kipwise (@kipwise_com) header
Kipwise (@kipwise_com) profile picture

The AI teammate for your Slack threads. It follows up, posts the status update, updates the record and writes it up, so nothing falls through the cracks.

Joined July 2016
Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)Photo by Kipwise (@kipwise_com)
IBM and Microsoft just announced an enterprise AI platform backed by more than 6,000 prebuilt agents and assets. 6,000. At some point, I wonder whether the challenge for enterprises will be figuring out which agents they actually need. We've spent years complaining about software sprawl. Agent sprawl could be considerably more entertaining.
12
3
27
1,264
McKinsey told Reuters that banks' routing of each query to a cheaper or pricier model fails once AI agents talk to each other, as tokens pile up. Track cost per finished task, and cap how many messages agents may trade on one. reuters.com/business/finance…
10
Small firms file incident reports, then the same thing happens again because nobody owned the fix. Our new guide on the Kipwise blog: a one-page form with an owner, a due date and a sign-off, plus a Monday check in Slack for open ones. kipwise.com/blog/incident-re…
6
A new hire's first week is one long list of questions, and they all land on the same person. Where do I set up payroll? How do I book time off? Most of it is written down. In our new one-minute film they ask in Slack, and Kipwise Agent answers with the page shown.
9
Kipwise retweeted
Vertical SaaS has historically been attractive because every industry has unique workflows that horizontal software doesn't handle particularly well. But what happens when building software around those workflows becomes dramatically cheaper? A distributor, manufacturer, or logistics company might have spent years looking for software that accommodates its specific processes. Now it can potentially build agents directly on top of its ERP, without buying another application. The irony is that the more specialized a workflow is, the more attractive custom development could become. Particularly when the software vendor is charging a premium precisely because there aren't many alternatives. _ We experienced some of this firsthand at our stealth co. Our first product analyzed company spending alongside software usage and other operational data to identify opportunities to save money. The more we tried to make it work horizontally across industries and company sizes, the more we realized how differently businesses operate, even when they're trying to solve what looks like the same problem. We eventually pivoted from building one product for thousands of companies to building custom AI workflows around how each company actually operates. It's an interesting position to arrive at after a month month trying to build the exact opposite.
I've been thinking about what happens to enterprise SaaS over the next few years. A few scenarios: 1. SaaS companies become the infrastructure for their own replacements. The optimistic scenario for incumbents is that they expose their products through APIs and MCP servers, allowing agents to operate their software on behalf of employees. But there's an interesting risk here: once the agent becomes the interface, the customer no longer needs much of what made the application valuable in the first place. The vendor retains the database and transaction engine, while the agent owns the workflow and user relationship. That's a very different position to negotiate pricing from. 2. Enterprises realize they've spent decades adapting themselves to software that was never built for them. Historically, buying an imperfect SaaS product was still cheaper than building and maintaining a custom one. That assumption is becoming less reliable. If a company can retain a handful of engineers to build applications around its actual processes, the economics of paying seven figures annually for software that requires endless workarounds start looking questionable. Especially when much of the functionality is only used by a fraction of the employees paying for seats. 3. The best SaaS businesses may become the worst businesses to sell AI to. If an incumbent automates 80% of the work its customers perform inside its product, it could destroy a substantial portion of its own seat revenue. The customer wants fewer people doing the work. The vendor wants more people paying to access the software. That creates a genuine incentive problem. Some incumbents will successfully move to usage or outcome-based pricing. Others may be financially better off slowing down the very automation their customers want. 4. The cost of building software falls, but the cost of owning it becomes the deciding factor. Generating an application is becoming cheap. Keeping it secure, integrating it with changing systems, maintaining permissions, handling exceptions, and supporting it for years is another matter. This could produce a strange outcome: companies build far more software internally, only to discover that maintaining hundreds of custom applications creates a new kind of technical debt. The winners may be the companies that figure out how to make custom software disposable rather than permanent. 5. Enterprise software stops being sold primarily to the people who use it. Today, a large part of enterprise software distribution depends on convincing employees and department heads that a product will improve how they work. But if agents increasingly choose which tools to call, where to retrieve information, and how to execute tasks, the buying criteria change. Interface quality, onboarding, and user experience become less important for certain categories. Reliability, permissions, interoperability, and cost per completed task become much more important. Some products with terrible interfaces but excellent underlying infrastructure could become more competitive, not less. _ A company could keep its core systems, replace much of the software employees interact with, build its own workflows, and reduce its seat count dramatically. And the software vendors could find themselves processing more transactions than ever while capturing less revenue per customer. That's the tension I find most interesting: AI could make enterprise software more useful while making large parts of the enterprise SaaS business model less valuable.
11
2
66
13,691
Uber says AI agents booked 500 of its 320 million trips in Halloween week. Customers are slow to send agents. Start agents on internal queues your team already works through by the same rules each week. gizmodo.com/the-ai-agents-ca…
12
Anyone can build an agent over a weekend. But the value in an agentic implementation goes well beyond building the agent itself. 1. It starts with the infrastructure. Connecting the systems, setting permissions, making sure somebody can see what the agent did and step in when something goes wrong. 2. There is the data. Where does it live? Is it current? Do two departments mean the same thing when they say “revenue”? An agent needs those questions answered to do useful work. 3. Change management needs to happen alongside the build. People need to understand what changes in their day-to-day work, what they’re still responsible for, and when to take over. Somebody has to work through that with the team. 4. You need evals. Define what a good outcome looks like before you build, test against real examples, and keep checking performance once the agent is in use. All of that needs time, budget and somebody responsible for it. Otherwise, you can ship a working agent and still leave the business doing things exactly as before.
41
32
4
230
47,108
Capgemini: 59% of large firms say AI spending ran over budget, by 24% on average, and many plan to move to cheaper open-weight models. That move happens one step at a time. Record which model each workflow step calls, so easy steps go first. reuters.com/business/most-la…
1
15
if your AI prototype only works half the time, try making it do less. you've got a "second brain" full of markdown files and saved notes. now you want a system you can actually rely on. the temptation is to ask the agent to build the whole production version in one shot, even when you don't know what that end state should look like. start by getting specific about one useful job it needs to do. ask your agent: "help me scope a smaller version of this. propose options and explain the tradeoffs. what can we leave out? what would i need to check before relying on it?" choose the narrow version. make clear what a good result looks like. try it on the work you actually need it for and work through the failures. a smaller scope gives you fewer behaviors to understand and check. once you've got that working reliably, add the next piece.
19
2
85
2,717
United will block AI agents from booking flights, partly because check-in makes passengers confirm their bags hold no banned items. Find the steps in your workflows where a person attests to something. An agent can fill that form, a human signs it. washingtonpost.com/travel/20…
19
Every Monday, you write the same list. Then you spend the morning chasing it. Say it once instead. In our one-minute film, Kipwise Agent in Slack takes the list: it checks what slipped, drafts the nudges, and waits for your OK before anything goes out.
1
17
Every week a CFO tells me the board wants AI in finance before year-end. Year-end is seven weeks away, and most of the finance teams we work with run on a zoo of systems, with Excel as the binding agent, because every acquisition brought its own books. You don't have to fix all of that first. Before an agent touches a number your auditor will see, make sure it has: > every figure traced back to the source system and the exact record behind it > reconciliation tolerances finance agreed on, so "close enough" means the same thing every month > cut-off rules written down, so a transaction on the 31st lands in the right period > intercompany eliminations checked across every entity, not just the big ones > the agent prepares the entry, a person posts it > a golden set built from last year's close, with the numbers your auditor already signed > a log of every adjustment it proposed, and who approved it > the model, prompts and mapping rules locked until the audit is done The LLM comes in last.
21
6
2
49
7,326
Databricks CEO Ali Ghodsi says he uses its Genie agent to surface problems his 12,000 employees are unlikely to bring to him. An agent can only report what got written down. Bad news that lives in DMs and hallway chats stays invisible to it. techcrunch.com/2026/11/11/da…
18
Your new assistant starts Monday. Tonight, you're deciding which passwords to send. Send none. Give them their own seat in each tool, with their own role. Our one-minute film then shows the weekly chores Kipwise Agent does in Slack, as you, asking first.
3
7 AI themes that CIOs must master: ➡️1. The AI agent reality check: agent autonomy is earned based on reliability, accountability and security. Three types of AI agents: personal, process (runs a defined workflow) and role (owns an outcome) ➡️2. The AI value gap is really an AI risk gap - security, cost and trustworthiness AI cost drivers compound as adoption grows. But cost risks will be solved. Trustworthiness and security and your forever risks. Policy is becoming code. Business context is a competitive moat. AI agents need precise instructions. ➡️3. The investment strategies of AI-first organizations - 4X more investments in AI for AI-first orgs vs AI-cautious or AI- opportunistic orgs. AI-first organizations invest narrowly - use cases save costs, improve experience and increase speed. AI-first companies reuse relentlessly. No one-offs. They spend 2X on foundations (data and context layer) than tools. ➡️4. The future of enterprise applications - the death of SaaS is false and a headless utility by 2032. SaaS is becoming a headless utility. SaaS is disappearing into the technology stack. SaaS is becoming very hard to replace. ➡️5. The AI shaped organization - everyone becomes a manger of tools and agents. We have a careless consumption of AI models and misuse of token spend. Managers as middleware are out; coaches and leaders are in. Incumbents can borrow some but not all AI-native processes. Tiny teams depend on high-agency employees. Smaller teams can produce more. Roles are compressing and expanding. Start with the work to be done. IT is customer zero. ➡️6. Volatility and sovereignty’s impact on the enterprise - we are in a trust recession. Super integrated systems is now an exposed. Optionality must be refunded to be real. Volatility is a reshaped advantage. ➡️7. Physical AI goes far beyond robots 🤖 - physical AI will make other forms of AI legacy by 2035. Autonomous vehicles, smart infrastructure and robots are all becoming smart. Orchestration of physical AI is costly. —Mary Mesaglio is the Chief of Research for CIOs and AI Leaders
14
8
1
24
10,181
PwC will train 85,000+ staff in critical thinking and judgment, since AI now does the groundwork where people used to learn it. Skepticism needs something to grip: have every AI summary show the source behind each figure. pwc.com/us/en/about-us/newsr…
10
You wrote the SOPs. Your team still asks you. The refund rule is on page twelve, in a folder, inside another folder, so they ask whoever wrote it. In our new one-minute film they ask in the Slack thread, and Kipwise Agent answers with the rule and the page it came from.
1
32
Gusto data via CNBC: small businesses that adopted AI had about 7% more staff a year later than similar firms that didn't. So build the owner's AI setup for the first hire to run. Prompts kept in one person's chat history can't be handed over. cnbc.com/2026/11/07/americas…
120
In an Anthropic study, AI agent teams beat single agents on business goals but produced less ethical solutions, partly by ignoring agents that raised concerns. When an agent flags a problem, route it to a person who can pause the run. cnbc.com/2026/11/08/ai-agent…
2
1
48
The Bank of England says AI makes UK firms less likely to replace staff who leave, yet productivity rose only 0.5% in three years. When a seat stays empty, track how long that person's old queue now waits. It shows whether automation took the work. theregister.com/ai-and-ml/20…
1
1
28