Zeer Pipe17 / intent / pipe17

Pipe17

51-200 employees·Seattle, Washington, United States·pipe17.com

Observed, not inferred · updated weekly

Buying intent

22 tracked signals | Top 15 topics are below | Marketing and Sales are carrying most of it.

21signals · 30 days
16topics tracked
58.33%attributed to a team

Attention by team

LinkedIn activity, by team

Where Pipe17's own people are actually spending their attention, by team, by topic. Bands run Low to High against the busiest pairing on this page, and each cell also shows how much of that team's own activity it represents.

Artificial Intelligence
AI Agent Software
Agentic AI System
Account Executive
DCL Logistics
Data Model
Marketing
High50% of team Marketing to Artificial Intelligence: High, 50% of this team's signals
Medium25% of team Marketing to AI Agent Software: Medium, 25% of this team's signals
Medium25% of team Marketing to Agentic AI System: Medium, 25% of this team's signals
Marketing to Account Executive: no signal
Marketing to DCL Logistics: no signal
Marketing to Data Model: no signal
Sales
Medium33% of team Sales to Artificial Intelligence: Medium, 33% of this team's signals
Medium33% of team Sales to AI Agent Software: Medium, 33% of this team's signals
Medium33% of team Sales to Agentic AI System: Medium, 33% of this team's signals
Sales to Account Executive: no signal
Sales to DCL Logistics: no signal
Sales to Data Model: no signal
Others
Medium20% of team Others to Artificial Intelligence: Medium, 20% of this team's signals
Medium20% of team Others to AI Agent Software: Medium, 20% of this team's signals
Others to Agentic AI System: no signal
Medium20% of team Others to Account Executive: Medium, 20% of this team's signals
Medium20% of team Others to DCL Logistics: Medium, 20% of this team's signals
Medium20% of team Others to Data Model: Medium, 20% of this team's signals
LowMediumHigh·  banded against the busiest pairing on this page

Topics being researched

30-day window

Every tracked topic, ranked by volume, not by our guess at what matters. Confidence is the classifier's own certainty that a signal belongs where we've filed it.

Artificial Intelligence
LinkedIn
High volume
96%
last
17d ago
AI Agent Software
LinkedIn
High volume
94%
last
17d ago
Agentic AI System
LinkedIn
Medium volume
98%
last
17d ago
Account Executive
LinkedIn
Low volume
98%
last
12d ago
DCL Logistics
LinkedIn
Low volume
98%
last
18d ago
Data Model
LinkedIn
Low volume
98%
last
11d ago
Purchase Order
LinkedIn
Low volume
98%
last
18d ago
Career Development
LinkedIn
Low volume
98%
last
24d ago
Software Development
LinkedIn
Low volume
94%
last
26d ago
Follow-Up
LinkedIn
Low volume
98%
last
17d ago
Go to Market
LinkedIn
Low volume
98%
last
23d ago
Systems Integration
LinkedIn
Low volume
98%
last
26d ago
Press Release
LinkedIn
Low volume
98%
last
11d ago
Instructional design (ID)
LinkedIn
Low volume
92%
last
18d ago
Platform.sh
LinkedIn
Low volume
78%
last
17d ago

Need intent for a specific topic or industry?

We track the full taxonomy across every account in the graph — including themes not shown on this page.

Talk to us

Who's active at Pipe17

verified title on file

Titles, seniority and topic straight from each person's own activity, with a LinkedIn link so you can check any of them yourself.

Others4 active
Executiveresearching Purchase Order
in
researching Instructional design (ID)
in
researching Account Executive
in
researching Software Development
in
Sales2 active
researching AI Agent Software
in
Directorresearching Platform.sh
in
Marketing2 active
VPresearching AI Agent Software
in
researching Software Developers
in

Primary products / business lines

LinkedIn company profile

Legacy order management was built for a world with fewer channels, simpler fulfillment, and more patience. That world is gone. Pipe17 is the AI-native Order Operations platform. One place to connect every selling channel, fulfillment provider, and ERP. One system to orchestrate, automate, and monitor every order. No integration spaghetti. No six-month implementation. No waiting on IT to flip a swi

Top accounts researching Pipe17

names withheld on the public page

These are companies whose own people brought up Pipe17 unprompted, not accounts we guessed might be interested. We can't yet tell an implementation partner from a genuine buyer here, names unlock along with the buyer profile below.

151,426 companies · 839,012 people are researching Artificial Intelligence

Pipe17's own team shows 4 signals on this topic. No one outside Pipe17 has been seen researching the company by name yet — so this is the market it sits in, not a list of its buyers.

  • AI Agent Software25,734 cos · 91,495 people
  • Agentic AI System32,339 cos · 130,942 people
Sign up to see which companies →

Buyer profile

company size · seniority

Company size and how senior the people involved are, the two things that decide whether this is a real deal. Competitor overlap isn't computed yet for this account.

Unlock the buyer profile

Company-size breakdown and buyer seniority mix for accounts researching Pipe17.

Company size breakdown
Buyer seniority mix

Sign up to learn more

Create an account to explore buyer insights as they become available.

Sign up to learn more →

Buying committee functions

Employee job titles (LinkedIn)

Sales — 2 people; Engineering — 2 people; Marketing — 2 people; Leadership — 1 person

What's been said

public posts mentioning Pipe17

Up to 9 public excerpts naming Pipe17 from the past 365 days, across LinkedIn, Reddit, X, YouTube and other public sources.

LinkedIn

As the Plaid of commerce, we at Pipe17 integrate with some of the most awful, legacy, abandoned IT systems on earth. Here's a sampling of some of the worst: - Direct database access: We literally connect to the database and manipulate the data directly. Decades old drivers - EDI over SFTP: EDI is written out to a file and then SFTP'd back and forth - AS/400 (green screens): Lots of flat files - FTP and SFTP'ing flat files: Arbitrary file formats, some random binary encodings, S/FTP'd arbitrarily - EDI VANs: Big clearing networks where EDI vendors swap data - Polling web hooks: Pushes notification that something has changed In many cases, documentation doesn't exist. Someone who's been there for 40 years knows everything and has to personally engage. Sometimes you literally have to go on-site and buy doughnuts and sit down in front of a whiteboard. Individual holes have to be put in firewalls to gain access to systems. Logistics and back office software is frozen in the 80's and 90's and often physically cannot be updated. These companies will go out of business decades in the future never having touched it. They can't. Many of the selling channels we connect up with (Walmart, Amazon, TikTok, Shopify, etc) have newer tech, but require that you: - Join their partner program - Negotiate a Data Protection Agreement (DPA) - Submit to their initial and then ongoing security audits - Pay ongoing fees for the right to connect into their systems - Police your customers to ensure they don't violate the platform's Terms of Service In short - connectivity is HARD and is the problem to solve as merchants deal with more selling channels, back office systems of record, and fulfillment channels. Our unique hub/spoke network abstracts away all the underlying complexity of connectivity, in the same way that Plaid does so for financial services. Plug in your API key and you're good to go! Connectivity is far more than wiring together two APIs. On top of our network, we've built AI-native enterprise order management. But the two are deliberately kept as separate layers. I firmly believe that AI will allow anyone to build an OMS and all the required features. There are already dozens of great OMSes on the market. What AI can't and won't solve is the connectivity problem, which is why our network-first approach to order management is so differentiated and is winning in the market.

May 2026
LinkedInPipe17

Today we launched pipe17.ai ! It's a detailed how to use Pipe17 guide but ONLY for your agents. It's pages and pages of well formatted markdown explaining what our product does and how to use it, including well structured Q&A. This information is previously spread across our .com (which like most websites is JavaScript-heavy and hard for agents to read) and our documentation. This concisely puts it all in one place, with clear markdown. Just copy / paste the instructions to your agent (from the "Human") tab and your agent will then have everything available from the "Agent" tab. (Thank to Mathias Biilmann Christensen who I shamelessly stole this idea from. He has long been talking about optimizing for "agent experience" and he is now proven to be right) https://pipe17.ai/

May 2026
LinkedInArtificial Intelligence

Smart move. Won't fix stockouts on its own. Unifying planning is real progress. Optimizing safety stock by store, region, and local demand is exactly the kind of problem AI is good at. At Lowe's footprint, the upside is huge. But planning is only half the system. The other half is execution: intelligent routing and sourcing, real-time inventory sync across channels, and exceptions resolved before a customer hits an empty shelf. That layer is fragmented at almost every retailer I talk to, even ones with sophisticated planning stacks. A better plan routed into a fragmented execution layer doesn't eliminate the stockout. It just diagnoses it faster. This is the gap we're closing at Pipe17 . Planning decides where inventory should go. Order Operations makes it actually happen. So the question isn't "do we have unified planning yet?" It's: what keeps breaking after the plan is right? Where's that gap hitting hardest for you? Routing, inventory accuracy, exceptions, or vendor execution? https://lnkd.in/g6aZXvav

Apr 2026
LinkedIn

A week ago, Anthropic released Claude Design which allows you to build entire presentations that adhere to a specific design system. Here I show in 11 minutes how simple it is to build deeply customized presentations using a fictional company - Chubbs Peterson's pro shop. This has quickly become the default way that we at Pipe17 are building presentations because they're dramatically better and faster than taking the gold pitch deck and lightly modifying it for each pitch. First, start by going to Claude where you'll use it to build an entire outline. Here's the prompt I used: --------------------------------- In an alternate universe, Chubbs Peterson from Happy Gilmore survived and opened his own pro shop. - Pro shop has gone viral and is now selling $250m annually through Shopify, TikTok, Amazon, Walmart, and EDI - They use Celigo to connect up their selling channels to NetSuite. Getting expensive and difficult to sell through new channels - They have an owned warehouse running Logiwa and also use Ryder as a 3PL I'm meeting with Chubbs himself tomorrow and would like you to build a slide deck I can present in 20 minutes. Chubbs is a golfer by training, so be sure to use corny golf analogies throughout. Claude Design will be building the presentation for me. Build a full markdown file with detailed instructions for content and visual design for each slide. I want to simply paste the output here into Claude Design Design notes for Claude Design: - Stick to the Pipe17 design system - Make the slides visually appealing - use lots of golf imagery. Avoid using too much text - Chubbs is a CEO - Ensure text isn't over the top of images, so that it's hard to read - Double check that the slides look great visually --------------------------------- In a normal pitch, Claude would have access to: - HubSpot / Salesforce data - Transcripts from previous meetings - Intel from various SI and ISV partners - Emails / Slack All of that data is then used to build the outline. You can get as detailed as you wish in building the outline. Make sure to use Opus 4.7 (the newest model) and turn on adaptive thinking so it thinks longer. As you can see in the video, it then spits out a full outline including detailed visual guidelines. Copy that outline and then paste it into Claude Design, selecting your company's design system. Your design system captures visual elements, color palettes, fonts, spacing, rounding of buttons - hundreds of visual design elements. Then you can see in the video how Claude Design actually builds the slides, while adhering to your company's design system. It's not perfect by any means. You'll want to swap out the clip art-type photos with real photos, and hit "Comment" to have it update the design in real time. Then export to .PPTX which you can then import into Google Slides. This has been on the market for exactly eight days. It's only going to get better from here!

Apr 2026

Leadership changes

Web / press

Pipe17 Raises Series A and Kelly Goetsch Joins as First COO; E-commerce logistics startup Pipe17 lands $15.5M and hires its first COO – GeekWire; After nearly five unforgettable years, I’m saying goodbye to my role as CMO at Pipe17 and beginning a new chapter with a soon to be announced stealth startup working in the AI …

M&A / partnership appetite

Web / press

Pipe17 Launches Industry-First MCP Server: Bring Order; We're excited to announce our investment in Pipe17, the order management platform built for how modern commerce actually operates.; Pipe17 Raises Series A and Kelly Goetsch Joins as First COO

Regulatory exposure

Web / press

Pipe17 Trust Center: Security & Compliance Docs; Pipe17 Launches SSO and Achieves SOC 2 Type II Certification; Pipe17 has filed a notice of an exempt offering of securities to raise $18,866,917.00 in New Equity Investment. - intelligence360

About this data. Pipe17 (pipe17.com). Department attribution is 58.33%; remaining activity is grouped under Others. Updated 2026-10-03T14:00:03.490Z.

Not yet computedPer-team narrative summaries