Linear #143: The Story of Augment, How To Launch AI Agents, 15 Quick Things To Finish The Year Out Strong
One vSaaS breakdown. One biz story. One 'how to'. In your inbox once a week.
Today’s newsletter is sponsored by Unit, the leading embedded finance platform for vertical SaaS.
Become the financial OS your customers rely on. With Unit’s Ready-to-Launch solutions, you can embed capital, banking, and bill pay in weeks - zero build required. Trusted by platforms like Bill.com, HoneyBook, and Homebase.
Learn more at unit.co
Alright, let’s get to it…
The Story of Augment:
Vertical AI For Logistics
I've been watching the logistics space for a while now, and I can tell you: what Augment just shipped is different. Really different.
While everyone else is building yet another TMS or trying to digitize freight matching, Harish Abbott and his team have quietly built something that actually solves the day-to-day operational nightmare that is modern logistics. They call it Augie, and it's the first AI teammate purpose-built for freight operations.
Here's why this matters—and why you should care.
The $2 Trillion Problem Nobody Talks About
Let me paint you a picture. Meet Christina. She's an operations manager at a freight company, and her day looks like this: 500+ emails, 40+ voicemails, endless phone calls, and jumping between a dozen different systems just to move one truckload of freight.
For every single load she books, Christina has to:
Send 10-15 emails
Make 4-5 phone calls
Log into multiple systems
Create dozens of entries
Download and upload countless documents
And trucks run 24/7. This never stops. Christina is drowning in busywork when she should be building relationships, negotiating better rates, and solving complex routing problems.
"There are frontline operators in over 200,000 shippers, 900,000 trucking companies and 20,000 trucking brokerages in America dealing every day with barrages of emails, phone calls, and data entry across a hot mess of fragmented software systems."
This is a $2 trillion industry being held back by stone-age operations.
From $2.1B Exit to Freight Revolution
Harish Abbott knows something about building transformative logistics companies. His last company, Deliverr, was acquired by Shopify for $2.1 billion in 2022.
But Harish's story starts way before Shopify. He cut his teeth at Amazon during the early days of Prime, learning how world-class logistics operations actually work. Then he co-founded Deliverr, which promised two-day shipping for any e-commerce merchant—a massive technical and operational challenge that required reinventing fulfillment from the ground up.
After the Shopify acquisition, most founders would retire to Napa. Harish went back to logistics. Why?
Because he saw the same operational chaos at freight companies that he'd solved at Deliverr, but 10x worse. And with generative AI, he finally had the tool to fix it.
Augie: The 24/7 Logistics Teammate
Here's what makes Augie different: it's not another dashboard or API integration. It's an actual AI teammate that can do the work.
Augie can read and write documents, respond to emails automatically, make and receive phone calls, log into existing systems, handle data entry and uploads, and learn company procedures instantly
The magic is in the details. Augie understands logistics—bills of lading, carrier requirements, load boards, tracking protocols. It doesn't just automate; it collaborates. When it needs help with a decision, it asks. When it learns something new, it remembers.
Christina doesn't need to learn new software or change her workflow. She just messages Augie on Slack or Teams like she would any other teammate.
This feels like the perfect AI Agent use case…
The GTM Playbook
Augment's go-to-market strategy is textbook enterprise software. They're starting with the largest freight companies—the ones with the most operational complexity and the biggest pain points.
The beauty of their approach: Augie doesn't require massive system overhauls or months of integration. It plugs into existing workflows and starts delivering value on day one.
Early customers are reporting dramatic efficiency gains. One operations manager described it as "having a personal productivity wizard by my side...feels like magic!"
The land-and-expand opportunity is massive. Start with operations teams, then expand to dispatch, customer service, accounting—anywhere there's repetitive, rule-based work that requires logistics expertise.
And they’ve raised $110M in 5 Months…
When a company raises $110 million in five months, you pay attention. When it's led by 8VC—a very smart enterprise investor—you really pay attention.
This isn't just about the money. It's about the signal. Smart investors see what we see: Augment isn't just building another SaaS tool. They're building the future of knowledge work in logistics.
The rapid follow-on funding round tells us two things:
Early traction is exceeding expectations
They're moving fast to capture market share before competitors wake up.
The Bigger Picture: Reinventing Work Itself
Here's what gets me excited about Augment: they're not just solving logistics. They're proving that AI can be a true collaborative partner, not just an automation tool.
The traditional approach to enterprise AI has been all wrong. Companies build "AI-powered" features that require users to learn new interfaces and change their workflows. Augment flipped the script: Augie adapts to how people already work.
This model—the AI teammate that truly collaborates—could work in any industry with complex, repetitive knowledge work. Legal, healthcare, financial services, manufacturing. Augment is building the playbook for the future of human-AI collaboration.
Why This Matters
Logistics is the invisible infrastructure that makes everything else possible.
When you order something online, when goods move from factories to stores, when the economy actually functions—it's all built on the backs of operators like Christina who are drowning in operational complexity.
Augment is giving these operators their lives back. And in doing so, they're making the entire supply chain more efficient, more reliable, and more human.
That's not just a business opportunity. It's a mission worth building.
Watch this space. Watch this company. Watch this founder.
Seeing that he has a $2B exit under his belt already, he is likely going to try to build this thing into a juggernaut…
Mini Playbook: Launching AI Agents At Your Vertical SaaS
Look, I'm tired of seeing vertical SaaS founders plug in AI chat bots. After implementing AI Agents across multiple verticals and companies the last ~6 months, here's how to do it, including what actually works—and what doesn't.
Start With Your Existing Workflows, Not AI
Stop asking "How can our customers use AI inside our platform?" Start asking "What repetitive work is killing our customers? What work are they spending an insane amount of time and money on?" I see founders jumping straight to ChatGPT integrations when they should be mapping their customers current operations first.
The sweet spot? Tasks that are:
High-frequency but low-complexity
Rule-based with clear decision trees
Currently eating 20+ hours/week of human time
Not life/death ie. if an AI hallcuinates your not going to lose a customer.
In healthcare SaaS, this might be prior authorization checks. In legal tech, contract clause extraction. In construction, permit status updates. You know your vertical's pain points better than any AI consultant.
The 6-Week Implementation Sprint
Week 1: Data Audit
After you’ve found a problem with a real ROI for your customers, it’s important to note that your vertical AI agent is only as good as your data. Period. Spend this week cleaning house:
Map every data source the agent will need
Check data quality (you'll be horrified)
Identify missing integrations
Set up proper data pipelines
Real Talk: I've seen $50K AI projects fail because someone assumed their CRM data was clean. It wasn't. Ever.
Week 2: Build the MVP
Start stupidly simple. Use existing tools like Zapier, Make.com, or even basic APIs before building custom infrastructure. Your first agent should handle ONE workflow end-to-end. Assign ONE person at your company to own this for the next few months. Someone good that you trust. One throat to choke.
Tech stack that actually works:
OpenAI API for text processing (yes, it's basic, but it's reliable)
Anthropic Claude for complex reasoning
ChatGPT Agent for basic data entry automation
Your existing database (don't build new ones yet)
Webhooks for real-time triggers
Week 3: Test With Real Users
Deploy to 3-5 friendly customers. Not your entire user base. Watch them use it. Most importantly—watch where they get confused. Make sure they/you can see a path to real ROI throughout this process. It doesn’t have to be crazy initially, but it needs to be clear that you are marching towards somethng real.
Week 4-6: Iterate On Feedback
Iterate the product accordingly with the ROI in mind. Remember the goal here is to eliminate some fraction of REAL W-2 time with AI. Stuff you couldn’t do previously with SaaS. AI is improving so fast that you should really start to see it improving significantly week over week.
Real world example:
When I deployed a data entry - the first week it got ~25% of the job done accurately. The second week it rose to ~50%. By week four, we were up to 90%. It was MIND BLOWING.
But we picked a really strong workflow that…
a) Folks were spending a crazy amount of time
b) Was very basic / straightforward manual data entry
c) Wasn’t life or death if the AI halucinated
The Mistakes That Will Kill You
Mistake #1: Building a chatbot. Your customers don't want another thing to talk to. They want work done automatically.
Mistake #2: Trying to replace humans completely with the initial launch. Build agents that make humans 10x more effective, not unemployed.
Mistake #3: Ignoring compliance. Healthcare, finance, legal—these verticals have rules. Your agent needs to follow them or you're toast.
Mistake #4: Over-engineering. You don't need a custom LLM. You probably don't need vector databases. Start with API calls and basic logic.
Pricing and Positioning
Here's where vertical SaaS founders have a massive advantage: you understand your customers' economics.
If your agent saves a law firm 10 hours/week at $300/hour, that's $156K annually. Price accordingly. Don't compete on features—compete on ROI.
Position it as "intelligent automation" not "AI." Your customers buy outcomes, not technology.
What's Actually Working Right Now
Document processing: OCR + LLM for data extraction
Customer communication: Automated responses based on ticket classification
Data enrichment: Filling incomplete records from multiple sources
Compliance monitoring: Flagging issues before they become problems
The Reality Check
Most AI agents aren't replacing entire job functions. They're handling the boring stuff so humans can focus on high-value work. That's still incredibly valuable.
Your competitive moat isn't the AI technology—it's your deep vertical knowledge. Use that. Build agents that understand your industry's specific language, regulations, and workflows.
Stop overthinking it. Start building. Your customers are waiting.
Next Steps: Pick ONE workflow. Map the current process. Build a basic automation. Ship it to five customers. Iterate based on their feedback. Repeat. Commercialize it. Go BIG with your launch.
Let me know how it goes…
Excited to hear how this worked for you!
15 Quick Things You Can Do NOW To Finish The Year Out Strong
We are coming up on the final few months of the year. Many of us are sweating our EOY numbers. I wanted to build out a list of 15 things you can do now to grow faster and ideally hit and/or exceed that number…
Do a mini customer conference — even if it’s in a cheap hotel or your friends office space. You can do this on a budget AND IT WILL accelerate your revenue.
Go on a road show — This ALWAYS WORKS. Set up 20-30 f2f meetings with existing or prospective customers. You will be amazed at how much additional ARR you can add.
Set up a tier bonus plan tied to your ARR goals — show me the incentive and i’ll show you the outcome. Too many founders don’t have bonus plans tied to their ARR north star. I PROMISE YOU what you will pay out will be worth it if everyone is rowing in the same direction.
Go visit your top 20 customers — Get on an airplane and take EVERY ONE of your top 20 customers out to a fancy dinner. Figure out what is working, what’s not, and ASK THEM what you can build for them that they need. Actually build said thing and watch the cross sell take effect.
Spin up a website for onboarding — Do you have a website for customers where they can see how long onboarding is taking? Where they can punch in the pain pre-implementation and the results post-implementation? They are busy and are not thinking about you, give them a website they can see everything on. Send them a link to it every 30, 60, or 90 days. PROVE YOUR ROI CONSISTENTLY and via documentation. Churn will decrease. I promise.
Dream up a new product offering with your 5 largest customers — Collaborate directly with them to brainstorm ideas that solve their biggest pain points. CO-CREATE it with them. Get them to commit to an annual number up front.
Launch a referral program with real incentives — Offer discounts, free months, or exclusive features for every successful referral. WORD OF MOUTH IS GOLD—make it easy for happy customers to bring in new ones, and you'll see your customer base grow exponentially without massive ad spend.
Optimize your pricing tiers — Review and tweak your pricing based on customer feedback and competitor analysis. INTRODUCE A MID-TIER OPTION if you don't have one—many founders miss how this can capture more market share and boost average revenue per user overnight. Think about raising prices. Chances are you probably can.
Host weekly webinars or AMAs — Share industry insights, product tips, or live Q&A sessions. THIS BUILDS TRUST AND AUTHORITY—prospects will convert faster when they see your expertise in action, and existing customers stay engaged, reducing churn. You have to do it religiously EVERY WEEK. Most companies do sporadic webinars. Doesn’t work.
Send gifts— A lot of great gifting platforms out there that always get people to respond. Send thoughtful gifts to your top 25 prospects.
Build a community forum, Facebook page for your industry, or Slack group — Create a space for customers to connect, share tips, and provide feedback. ENGAGEMENT BREEDS LOYALTY—watch as users solve each other's problems, freeing up your support team and uncovering new feature ideas organically.
Use AI photos of people your trying to get in touch with— Create a funny/entertaining photo of a prospect using AI that your trying to get in touch with. Folks typically reply to this!
Partner with complementary businesses — Create a simple partnership with another company that has a good product for your customer base and vice versa. Set up a webinar with your customers and do a webinar with theirs.
Trade introductions — Intro a complimentary company to 10 of your customers, have them do the same thing.
Drop a press release on every new customer you sign up— Too few people do this and really compounds and can create momentum in your industry.
Just a few ideas for you all to get the wheels turning a bit~!
Have a product or service that would be great for our audience of vertical SaaS founders/operators/investors? Reply to this email or shoot us a note at ls@lukesophinos.com