Where “imagine if” gets to work. We've helped 500+ companies (like @NotionHQ, @Roblox, @Uber, @Square) take a straighter path from idea to product-market fit.
The mechanics of sales may be automated away, but that was never the hard part.
We want to find and train the next generation of sellers who excel at the uniquely human craft of sales (and aggressively use AI for all the rest), so we’re launching Sales Combine.
Spend two days in SF for FREE training with extremely experienced sales leaders, and then we’ll introduce you directly to companies like @NotionHQ, @fal, @togethercompute, @p0, @reductoai, @Gumloop, @clay, & many others that are hiring new sellers.
Apply by 10/3: firstround.com/sales-combine
Manifest Law has quietly lined up so many of the dynamics that get us excited about AI-native services:
1. Significant recurring revenue from large enterprises.
2. Fragmented incumbents.
3. Verifiable output (not subjective.)
4. Regulatory moat.
5. Right balance of complexity and simplicity in the work.
6. Lots of admin around the core unit of work.
7. Potential for multiple lines of business early on consumed by the same end customer.
I sat down with Dan Mishin to explore what they’ve figured out in building an AI-native services business.
Timestamps:
00:00 Introduction
01:27 From selling antiques to a hostel at 11
05:36 Signing a 99-year Moscow lease at 18
08:17 Why AI-native services won't repeat the 2010s bust
11:06 How price collapse helps Manifest beat 120-year-old firms
12:47 Why Manifest goes all-in on immigration only
15:51 Fixing quality at scale with an AI case health tracker
18:26 How a 20-hour immigration case became 20 minutes
21:09 Earning the right to change how lawyers work
25:09 What really convinces immigration lawyers to join Manifest
29:09 Rebuilding the product as visa rules change weekly
36:26 Why Manifest killed, then revived, North Star planning
38:07 Selling AI in person with 350 events a year
41:39 Why every AI category leader has cracked marketing
46:23 Why brand is the most underrated moat in tech
48:56 Reinventing yourself as CEO in the age of AI
53:14 Why first-time founders micromanage, and how Dan stopped
I’ve now written a fair number of “@fal raised $$$” posts since backing their seed @firstround in 2022. Today it’s a two-for-one, with @sequoia leading the Series E and @BessemerVP leading the Series F (with @nvidia continuing to invest) for a total of $1.65B.
The common theme across every milestone and update is that the fal team’s greatest strength is their relentless pace. They obsess over finding new ways to make image, video and audio models better, faster and more useful to more people.
@burkaygur and @gorkem are particularly excellent at continually pushing on the bounds of what’s possible and staying far ahead of the market.
They’re simultaneously managing the complex game of building out infra to run models at incredible speeds and scale, while also doing their own research to move the frontier forward.
This year they dove into post-training already-great models to be even better and faster, and the results with H3 Max and FLUX 3 Max have been amazing.
Very proud to have partnered with this team from the very beginning, and can’t wait for everything they have cooking for the coming year.
Most people say the best venture investment decisions are the most contrarian ones, where one investor pounds the table and no one else sees it. That wasn’t the case when we met the @fal team for their seed round. 100% of our team wanted to invest.
What the @fal team has done since then is astonishing. Can’t wait to see what they do with a lot more fuel in the tank.
Today, we’re announcing we raised our Series E & F rounds for a combined total of $1.65B. Our Series E was led by @sequoia and our Series F was led by @BessemerVP with continued investment from @nvidia.
We’re excited to push the boundaries of what’s possible even more next year and this funding will enable the team to build the way we want to.
our team has been heads down building for years and I'm just so grateful for the response to Phonic-1's launch yesterday
I love my team. they built a model that's..
#1 on AA's speech-to-speech index
#1 on AA's 𝜏-Voice (agentic performance)
#1 on Sierra's τ³-Voice
#1 on Cekura's s2s benchmark
#1 on Veris's VAmoS Pro Bench
#2 on Bluejay's MIVAS s2s benchmark
we literally just dropped seven figures on a new compute cluster to support the load
we can't wait to continue pushing the frontier for our customers!! try it out and let me know what you think
We raised a $150M Series C at a $3B valuation, led by @sequoia. Our third round in 12 months, bringing our total funding to $277M.
Our team has grown 10x in the last 12 months. We now serve Fox Corporation, WPP, Live Nation, Lucid Motors, and Ramp, and many of them are fully replacing legacy platforms with @getserval.
AI coding tools have transformed how engineers work. Serval brings that same power to every team, so anyone can build governed automations across the systems they already use.
We’re hiring 500+ people over the next year and expanding in San Francisco, NYC, London, Austin, and beyond.
If you’re down to take on legacy enterprise systems and transform how teams work, my cofounder @AlexAtServal and I would love to meet you.
It’s Serval now.
Jake Stauch @jakeserval Oct 29 Article We raised $150M to make everyone a builder. Engineers got super intelligence first. AI coding tools have transformed the way engineering teams work. They’re building faster, shipping more, and tackling the problems that used to be out of reach.
You can assign your investors tickets, too. Congrats to @jakeserval and the @getserval team on today’s $150M Series C led by @sequoia at a $3B valuation.
Proud to be your partners since literal day 1.
Introducing Phonic-1, the most advanced speech-to-speech model for agentic tasks — from @phonic_co, the speech-to-speech model company.
We’ve had a breakout year and raised a $27.5M Series A led by @firstround with participation from @Lux_Capital, @felicis, and Definition.
Today, Seneca announces that @SBCOUNTYFIRE has become the first agency in California to acquire a @Seneca_Systems Strike Team of five aircraft plus a Mobile Air Operations Base.
This is the second fire department acquisition of an autonomous aerial fire suppression system in the world. It will being a new set of capabilities to our nation's largest county -- and one that regularly sees nearly 2,000 yearly vegetation fires -- early in 2027.
It also will get the world a little closer to the moonshot goal of @SBCFireChief Dan Munsey: instant response to keep fires at 100 square feet. Fire grows exponentially. Detection has improved. Autonomous aerial response is the clear next step. We are extremely to partner with the exceptional men and women of San Bernardino County Fire to push the boundaries of what is possible.
San Bernardino County Fire will also put the aircraft into use for fuels management and prescribed fire, cargo hauling, night operations, mop up, and other mission types that will keep firefighters and communities safer.
California's wildfire threat is existential and well documented. It is a true privilege to work with the heroes and leaders in one of the nation's largest and most active districts to turn the tide.
A special thanks to the exceptional broader team in SB for the hard work to make this possible: Robotics PM Jeff Alexy, BC Chris Bowden, @Chief_Serrano, @billclerico, Captain Shawn Millerick and to Chris Anthony for making the introduction to Chief Munsey back before both Seneca and FireWERX!
**On a personal note, Dan and San Bernardino was my second call when learning about the fire industry before Seneca. The department staff was generous with their time and wisdom expecting nothing in return. I feel extraordinarily grateful to now call those same individuals friends and partners
“It's better to be in the hottest category in tech than the coldest one, right?”
That’s what @jgreze, co-founder and CEO of @TownAI, said in our recent conversation. He’d been building an AI Assistant for nearly a year — then within two months, Meta, OpenAI and xAI all launched agents.
Most people assume that kind of competition is death for a startup. JDG sees it differently. He thinks the big players have distribution but no real network effects, and that incumbents look less scary than ever.
TIMESTAMPS:
01:28 Competing in the AI Assistant race
03:23 Why people will run two AI Assistants
07:01 Why personal AI hits a ceiling and work AI doesn't
09:08 What made always-on agents possible now
15:54 Building on labs that compete with you
17:57 Pricing tokens instead of outcomes
22:52 How Town's users drive new features
25:08 Why Town is actually an AI onboarding platform
29:13 The one-agent-per-person debate
32:01 Designing an agent people can trust
35:27 How multiplayer changes privacy
44:33 Why AI should choose what you share with coworkers
Our first Sales Combine cohort includes:
A chef
A beer sales rep
Someone who runs his family’s furniture store
A former talent agent from LA
We want to open the door to economic opportunity in technology for more people while giving our companies every advantage they can get.
This new effort is going to be a win-win.
The mechanics of sales may be automated away, but that was never the hard part.
We want to find and train the next generation of sellers who excel at the uniquely human craft of sales (and aggressively use AI for all the rest), so we’re launching Sales Combine.
Spend two days in SF for FREE training with extremely experienced sales leaders, and then we’ll introduce you directly to companies like @NotionHQ, @fal, @togethercompute, @p0, @reductoai, @Gumloop, @clay, & many others that are hiring new sellers.
Apply by 10/3: firstround.com/sales-combine
I’m excited to share that Mach9 has closed a $34M Series A co-led by @firstround and @BessemerVP , with participation from @Prudence_VC and @QuietCapital .
During the first week of my freshman year at Carnegie Mellon, I walked into Red Whittaker’s office wanting to build field robots. Red handed me blueprints of a nuclear waste storage tunnel and told me to turn them into a CAD model, simulate it, and then build a robot to actually inspect it.
Our robots and software helped infrastructure owners and engineers understand what was there and plan what came next. The problems we worked on had real consequences for people and their communities.
That experience shaped why we’re building Mach9.
So many of the smartest people in the world are building software to serve better ads or AI to optimize a sales process, while entire industries that keep the world running still rely on tools from a generation ago.
Civil engineering is one of them. While civil engineering teams are using tools like LIDAR that make collecting data on their projects super fast, turning that raw data into usable CAD drawings and designs still takes months of manual drafting. And there are simply not enough people to do the work.
At Mach9, we’re building AI CAD software to expand the world’s capacity to build and maintain the infrastructure people depend on. Our software turns reality capture data into centimeter-accurate CAD drawings, helping surveyors and civil engineers deliver projects faster.
We’ve brought together an exceptional team from SpaceX, Tesla, Apple, and NASA to tackle hard technical problems in an industry where better tools can have a lasting impact. We’ve already helped nearly 100 organizations deliver over 2,500 projects spanning 100,000 miles of infrastructure in all 50 states. Behind those numbers are real problems we’re helping customers solve:
- On I-8 near San Diego, Caltrans is using Mach9 to turn scans into an accurate inventory of road safety assets, helping their team plan upgrades to make California’s roads safer.
- Fibersmith is using Mach9 to help design fiber networks that bring high-speed internet to more than 120,000 homes and businesses in some of Louisiana’s most underserved communities.
I’m so proud of what our team has built, and grateful to the customers and partners who have shaped Mach9 along the way.
Whether you’re looking for better tools for your next project or want to help us build them, we’d love to hear from you. Join us in bringing civil engineering into the AI era
You’ve likely seen surveyors standing on the street with their yellow vests and tripods. These days, a lot of that work is done faster with LiDAR, but turning those scans into engineering drawings is still a surprisingly slow process. It requires weeks/months of manually tracing every curb, paint line, telephone pole, and sign. @Mach9Geo is solving this bottleneck for civil engineering teams, and we’re proud to be their partners at @FirstRound, co-leading their $34M Series A with @BessemerVP.
Their AI CAD software turns 3D scans of roads, utilities and construction sites into draft drawings for surveyors and engineers. Work that used to take months can happen in days, giving infrastructure teams the capacity to deliver at the pace we need (which is especially helpful given how hard it is to hire surveyors right now).
Teaching machines to read the physical world is one of the hardest problems in AI, but @alexbaikovitz, Zach Sussman, Haowen Shi, and Michael Mong are the ones to do it. They’re a world-class team with robotics roots at Carnegie Mellon and experience spanning SpaceX, Tesla, NASA, Apple and GoogleX.
And they’ve made tons of progress here (both technical and commercial). Mach9 has now been used across more than 2,500 projects in all 50 states for customers ranging from civil engineering firms to transportation agencies (such as Caltrans, Minnesota DOT and the Maryland State Highway Administration).
Whether it’s bringing high-speed internet to rural Louisiana, increasing pedestrian safety around schools in San Francisco, or road safety upgrades on Interstate 8 in San Diego, these are meaningful projects that move our country forward.
100K miles down, millions more to go.