Sabitlenmiş Tweet
Supply chain tech is a $30B+ industry.
And your team is still processing PO changes via email.
Here's the uncomfortable math nobody in this industry wants to talk about: 🧵
English
Ariel Palones
226 posts
Ariel Palones
@apalones
Supply chain tech spent $50B on dashboards. Execution is still done in email. Building the fix. | Co-Founder & CEO at Emerix
Katılım Ocak 2017
83 Takip Edilen77 Takipçiler
באיזה שעה אנשי פיקוד העורף מתכוונים להודיע על חזרה לשגרה?
עברית
@alonlif אולי נקח אותם כאאוטסורס לתע״ש ורפאל. עשו עבודה יפה על המלאים בחצי שנה.
עברית
אני לא בטוח, אבל יש סיכוי שחיזבאללה לא הושמד כמו שאמרו לנו
בצה״ל הציגו נתונים חדשים לממשלה על חיזבאללה: ״הארגון עדיין מחזיק בלמעלה מ-10,000 טילי נ״ט בטווח של כ-8 ק״מ בלבד מהגבול, ולמעלה מ-30,000 טילים ורקטות לטווחים שונים ועוד אלפים רבים של מרגמות ופצמרים - וקרוב ל2000 כטבמים.״
עברית
@feross Me, an AI supply chain startup CEO, clicking on "supply chain attack" trending:
Oh cool maybe people finally care about..…never mind it's npm again
English
🚨 CRITICAL: Active supply chain attack on axios -- one of npm's most depended-on packages.
The latest axios@1.14.1 now pulls in plain-crypto-js@4.2.1, a package that did not exist before today. This is a live compromise.
This is textbook supply chain installer malware. axios has 100M+ weekly downloads. Every npm install pulling the latest version is potentially compromised right now.
Socket AI analysis confirms this is malware. plain-crypto-js is an obfuscated dropper/loader that:
• Deobfuscates embedded payloads and operational strings at runtime
• Dynamically loads fs, os, and execSync to evade static analysis
• Executes decoded shell commands
• Stages and copies payload files into OS temp and Windows ProgramData directories
• Deletes and renames artifacts post-execution to destroy forensic evidence
If you use axios, pin your version immediately and audit your lockfiles. Do not upgrade.
English
"Most agents in production are RAG pipelines with a for-loop" - this.
The missing layer is business judgment. Can the agent validate an action against your specific rules before executing? Can it learn from corrections? Does it keep a full audit trail?
Without that, it's a demo, not a deployment.
English
What is Agentic AI?
Let's break it down (a thread🧵)
English
Every purchase order gets revised an average of 3 times.
Each revision touches 2-3 people and 2–3 systems.
For a company processing 500 POs/week, that's 4,500+ manual touchpoints per month.
Now calculate the error rate at scale.
That's where your 2% of your revenue goes.
English
@nicbstme Already happening in supply chain. Our AI agents process PO changes that come from a retailer’s automated system. Neither side is human anymore. The humans just approve the exceptions.
English
Exactly what is hard right now in product management is spending my day asking: How can I best serve my customer's agent?
Which is a totally different question to how can I best serve my customer.
There is an AI in between and the AI is now the gatekeeper to the relationship with the human customer.
Rohan Paul@rohanpaul_ai
Andrej Karpathy: "the industry just has to reconfigure in so many ways, like the customer is not the human anymore, it's agents who are acting on behalf of humans. And this refactoring will be probably substantial in the space."
English
Spot on. The part most people underestimate: the unstructured data isn’t just documents and emails.
It’s the corrections your team makes every day that never get written down. The override that becomes a rule. The workaround that becomes a process.
Agents that can capture and learn from that, not just read static data, are the ones that actually stick.
English
There is a huge opportunity for resourceful and entrepreneurial talent within organizations to go in and reimagine workflows for a world of agents.
The way you automate work with agents requires real work. It means setting up unstructured data in a way agents can easily access, learning the workflow and processes and creating skills or plans for agents to leverage, connecting disparate systems together, and likely changing the process itself to support getting the agents the need to do much of the work. Then you have to design where humans will play a role to oversee the workflows, how you validate the work, and so on.
Most of the gains you see from coding don’t take this level of effort because the agent knows more, it gets context more easily, and the users are technically. But for the rest of knowledge work there’s no way around this; there’s really no way to shortcut any of this work. It has to be done by a person or people on the team.
You will see a huge growth of roles within enterprises, and people that specialize in this will be hugely valuable in the economy. Great way for early career folks to make a huge dent quickly as well.
Alex Lieberman@businessbarista
Contrarian take: there's never been a better time for early career professionals to command the attention of leaders in their company. How? Become the AI Guru in your company. Every co has a few of them & they instantly earn visibility with the c-suite... If you're an SDR, map out the inbound SDR process end-to-end in your org and build an AI agent on Vercel that automates several/all steps. If you're an engineer, evangelize Claude Code & lead enablement workshops, helping other technical folks understand everything from hooks to worktrees to claude[.]md. If you're a growth marketer, build a paid media workflow that mines creative ideas from reddit, spins up 100s of ads using Nano Banana, sets up the campaign on Meta, and drives traffic to custom landing pages, so you can test ad spend far more efficiently at scale. You could be the most junior person in your company, but if you're truly viewed as the AI Guru, you are wildly valuable & have way more leverage than you think.
English
RPA in supply chain was a $10B industry experiment.
The conclusion: supply chain processes are not predictable.
Carrier emails change format. PO revisions contain natural language exceptions. Vendor portals update their UI.
Every RPA bot that breaks is proof that rule-based automation was the wrong answer.
English
@andrewchen It learns from your team’s corrections and tribal knowledge, not just training data.
Every override becomes a rule. Every exception becomes memory.
Fake AI resets every session. AI-native gets smarter every week.
English
how to tell the difference between AI-native products versus when AI is bolted on after the fact...
fake AI products:
- main AI feature is an AI button with sparkle icons
- chat pane where you can ask LLM questions
- no memory/personalization beyond one chat
- users try it once and go back to using the app the "normal" way
- AI is optional not essential to the product working
AI native products:
- you can spend $100 or $1000 via tokens as you use the product
- it gets substantially better every 6 months as base models improve
- core workflow is impossible without AI, not just enhanced by it
- creates behavior change when users try it
what else should be on this list?
English
@DeepLearningAI Exactly. In supply chain, the broken process is:
Carrier sends email → human reads it → human updates spreadsheet → human updates ERP → human emails warehouse
Adding AI to step 1 doesn't help. You need AI that replaces steps 1 through 5 with one governed workflow.
English
CEOs are wasting millions on AI. Here's why. 💸
CEOs are pouring millions into AI, but simply swapping a human for an AI agent in the middle of a broken process won't change your business. True AI transformation requires an end-to-end workflow redesign. Don't just use AI to make the old way faster—use it to build a completely new experience.
Ready to build AI workflows that actually move the needle? @DeepLearningAI offers free short courses to help you master agentic workflows and the latest AI tools.
Start learning today: hubs.la/Q048nRwq0
Subscribe to The Batch newsletter for weekly AI insights: hubs.la/Q048nVvj0
English
$901,600 recovered. 4 minutes. Two out-of-stock items.
A $3.2M purchase order came in. SAP flagged two lines as out of stock.
Normal day: emails flying, calls to the warehouse, someone pulling up a spreadsheet to figure out alternatives.
Instead: an AI agent found alternatives, matched on product line and availability, sent a change request to the customer, got approval, updated SAP.
The team's job? Review and approve.
Not a projection. This actually happened.
English
@Alfred_Lin The supply chain version of this:
Adoption = AI that reads emails faster
Advantage = AI that reads the email, validates the PO change against your business rules, updates the ERP, and notifies the warehouse - autonomously
That gap is judgment. And it's where all the value lives.
English
Everyone will adopt AI. Everyone will build more, ship faster, and deploy agents. That is table stakes. The founders who win will be the ones who make better choices, which requires judgment. (And if you're in the bottom left quadrant, get out quick.)
Alfred Lin@Alfred_Lin
English
@lennysan Hiring AI PM for @emerixai in Tel Aviv
We are building The AI execution layer for supply chain
Apply here: linkedin.com/jobs/view/4386…
English
Who’s hiring PMs right now?
Reply with the role, company, location.
English
@mattturck 2027: asking why the AI agents they sold you still can’t process a PO change without a human in a spreadsheet
English
What VCs do on company boards, an evolution:
2016: governance, guidance, support
2021: cheerleaders
2026: Anthropic and OpenAI sales representatives
English
Agree, but supply chain needs more than AI-fluent people. It needs AI-fluent systems.
Right now the smartest supply chain hire in the world still spends 70% of their day on data entry because the systems around them can't reason or act.
Fluency without execution infrastructure is just a faster human doing the same manual work.
English
"I would hire the one who is expert in using AI." - Jensen Huang, CEO of NVIDIA
@lexfridman asks about the future of hiring across every role, accountants, lawyers, marketers, supply chain, customer service.
Jensen's take: AI fluency is now required. For every department. No exceptions.
English
@awscloud 15-20 hours/week is one workflow.
Most supply chain teams are losing 10,000+ hours/year across dozens of workflows - all running through email, spreadsheets, and manual ERP updates.
The unlock isn't no-code. It's AI that understands your business context and actually executes.
English
What if your team could build & automate their operations—without writing a single line of code?
Amazon's Mexico Supply Chain team did exactly that with Amazon Quick, reducing weekly manual work from 15-20 hours to minutes.
By leveraging Quick Flows, tailored Spaces & custom agents, they've empowered teams to work smarter.
The impact? Improved delivery speed tracking & more time to innovate.
English
Talked to a $1.5B beauty brand last month.
They have 4 people whose full-time job is copying shipment data from carrier emails into their ERP.
Four people. ~$400K/year in fully loaded cost.
To copy and paste.
That's not a headcount problem. It's an architecture problem.
English
אל-על בתגובה: חובבנים. אצלנו תוכלו לשלם על שני מושבים באיקונומי ולישון במיטה זוגית קינג-סייז בדיוק כמו בבית! כי זה בבית. כי אתם נשארים בבית שלכם יא חמארים מה לא הבנתם?
United Airlines@united
The entire row is alllllll yours. Welcome to United Relax Row, three adjacent United Economy seats with adjustable leg rests that can each be raised or lowered to create a cozy lie-flat space for stretching out... You'll also get a mattress pad, blanket and two pillows. If you’re traveling with kids, a plushie too! United Relax Row will be available starting next year on more than 200 of our 787s and 777s, each with up to 12 of these brand-new rows. united.com/Elevated
עברית
Control towers are the most expensive screenshots in the enterprise.
You spent $2M so your VP can look at a dashboard that says "shipment delayed."
Cool.
Who's rescheduling the receiving appointment? Who's updating the PO? Who's telling the warehouse?
That's still a person. In email. Manually.
English
There's a third camp he's missing: Value Recovery.
In supply chain, companies lose 2% of revenue every year to manual execution - stockouts, chargebacks, OTIF fines, expediting.
AI doesn't create new revenue or cut costs. It stops the money that's already leaking out the back door.
That's not preservation or creation. It's collection.
English
AI will split companies into two camps.
Value Preservation — using AI to cut costs.
Value Creation — using AI to grow revenue.
Value Creation companies will be the ones hiring more people because of AI. Which kind of company do you work at?
English