Founders In Motion  /  Episodes  /  Ep 23
Episode 23 · AI · Future of Work · Seed

The 25-Year-Old Who Raised $6M From Early Facebook Investors

Released: 26/02/2026 Guest: Finnlay Morcombe, Co-founder, Fluency
In one paragraph: what's this episode about?

At 25, Finnlay Morcombe built Fluency — a platform that maps how work actually happens in Fortune 500 organizations — raised 6 million from Accel within weeks of landing in the US, and hard pivoted into the product within one week of going to market.

Answered by Finnlay Morcombe, Fluency — interviewed by Thea Ngo.

How Finnlay Morcombe did it: The 25-Year-Old Who Raised $6M From Early Facebook Investors

Most 25-year-olds are figuring out their first job. Finnlay Morcombe built Fluency, a platform that maps how work actually happens in Fortune 500 organizations. He and his co-founder Ollie are self-taught devs from Melbourne, and by the time of this conversation Fluency had 15 million in pipeline and Fortune 100 partners.

The raise almost didn't happen on purpose. Finnlay says they "weren't really sure we were going to raise at the beginning" — they had enough traction that a raise "wasn't imperative." So they wrote "a literal two pager," with "no deck no data room," sent it to about 10 funds, and ended up speaking to "18 funds or less total for the whole thing." What worked, in his telling: he had already moved to the US regardless of the outcome, the attitude that "we have enough capital in the bank anyway and we're going to get it done," real traction on the new product, and the fact that what they were building "perfectly hit a thesis" Accel had internally. Diligence was "12 background calls pretty much just on myself" — they even called his brother, a chippy, while he was on site with "hammers and stuff going in the background."

Fluency itself gets deployed onto everyone's laptop inside an organization. It takes screenshots, scrapes system logs, captures network logs, and from all of it builds what Finnlay calls a "work ontology" — a queryable graph of nodes and edges that is "non-monotonic," meaning "the present can affect the past." There are three use cases today: tell businesses what their processes actually are, tell them how to improve them (usually through automation), and measure the ROI on change itself. He stresses it is all done "in aggregate" and that "we don't allow individual employee monitoring or productivity indexing." Where it's heading is foundational models — not large language models but "world models" that predict "the next most likely event," with a "sovereign world model per customer." One of their investors "heads up AI at Princeton." His timeline: 1.5 in "8 to 12 weeks," act two in "9 months."

This is a company built on pivots. Before Fluency's current form, they were doing SOP creation — and "didn't feel the product market fit." Selling it was "like Copus pushing the ball up the hill." When they finally went to market with the work-mapping version, the read came in fast: "within one week we're like okay we're hard pivoting because the reception was completely different." The old documentation product is now, by his estimate, "10 basis points like .1% of the new product."

The conviction underneath it all is that they'd "rather company die trying" than "be a $50 million company" — they'd "rather spend all of our money and die at 100 million" than play it safe. Finnlay misses his family, his brother, his family's pets, his girlfriend, and his friends back home; the number-one cost of building Fluency, he says, is "time with family." But work, for him and the team, "is one of our hobbies." Asked what he'd do differently going back to the 2023 Swinber accelerator days, he lands on one word: "agency is so powerful." Just try and do — "what's going to happen?"

What you'll hear

  • The accidental raise — how a "literal two pager" sent to about 10 funds became a 6 million round from Accel before he was even fully moved in
  • What Fluency actually does — laptops, screenshots, system and network logs, and a three-layer "work ontology" of how work gets done
  • World models, not LLMs — why the next act is a "sovereign world model per customer" that predicts "what's going to happen next"
  • The one-week hard pivot — going to market with SOP creation, feeling no product-market fit, and switching within a week
  • Tall poppy syndrome — the three places Finnlay actually sees it in Australia, and why he doesn't really care
  • Die trying — why they'd rather spend all their money and die at 100 million than be a 50 million company
  • The cost of building — skateboarding, mountain biking, a wider friend circle, and most of all time with family

Key claims from this episode

6 million
Seed round Finnlay raised from Accel, before he'd been in the US more than a few weeks
15 million
In pipeline, with Fortune 100 partners, built by self-taught devs from Melbourne
18 funds
Or less, total, that he spoke to across the whole raise — off a "literal two pager"
.1%
The old SOP documentation product as a share of the new product's capability and value

Chapters

00:00
Cold open"agency is so powerful, what's going to happen?"
00:40
The 10,000-subscriber goal
00:58
The raise6 million from Accel within weeks of landing in the US
01:57
Advice for Australian founders raising in the USthe two misinterpretations
04:39
Do US investors discount Australian revenue?
05:31
What Fluency actually doesthe work ontology
09:08
Where Fluency is headingworld models, not LLMs
12:23
Is agentic AI just RPA on the cloud?the hot take he no longer holds
14:51
The pivot from SOP creationand the one-week hard pivot
16:27
Being in your early 20s selling to large enterprises
17:44
Tall poppy syndromeMelbourne to SF
20:05
The weight of a 6 million raise, and "die trying"
21:16
Loneliness, family, and what building has cost
22:46
What he'd do differentlyagency

Quotes from this episode

we went to market with it and within one week we're like okay we're hard pivoting because the reception was completely different
— Finnlay Morcombe, on the one-week hard pivot (15:48) we would rather spend all of our money and die at 100 million
— Finnlay Morcombe, on why they would rather die trying (20:46) what we were building perfectly hit a thesis they had internally
— Finnlay Morcombe, on why Accel led the round (02:59) agency is so powerful to just like do
— Finnlay Morcombe, on the one thing he'd do differently (22:46) we don't allow individual employee monitoring or productivity indexing
— Finnlay Morcombe, on how Fluency measures work (08:41)

Themes Finnlay returns to

  • Die trying over a safe exit — they'd "rather company die trying" and "spend all of our money and die at 100 million" than be a "$50 million company"
  • Agency — the through-line he keeps returning to: "agency is so powerful," you can "be super ambitious from day one and you can just do things"
  • Pivot on the signal — feeling no product-market fit on SOP creation, then hard pivoting "within one week" when the reception flipped
  • Work as a hobby — "we're lucky that I guess like work is one of our hobbies," even at "14 hour days you know six days a week"
  • Mapping reality, not surveilling people — a "work ontology" built "in aggregate," with no "individual employee monitoring or productivity indexing"
Full transcript ~5,500 words
This is an auto-generated transcript, lightly edited for readability. Timestamps reference the audio version. If you spot an error, let us know.

00:00:00.080 Most 25-year-olds are figuring out their 00:00:01.920 first job. Finnlay Morcombe built Fluency, 00:00:03.929 [music] a platform that maps how work 00:00:06.080 actually happens in Fortune 500 00:00:07.919 organizations. 00:00:08.720 >> We went to market with it and within one 00:00:10.240 week we're like, "Okay, we're hard 00:00:11.120 people." 00:00:11.519 >> And convince investors behind Facebook 00:00:13.840 to lead a $6 million seed round [music] 00:00:16.079 before he even unpacked in San 00:00:18.320 Francisco. 00:00:18.880 >> We would rather company die trying. 00:00:20.400 >> Yeah. 00:00:20.880 >> Than be like a $50 million 00:00:22.240 >> fluency. [music] 00:00:23.199 15 million in pipeline. Fortune 100 00:00:25.600 partners built by self-taught deaths 00:00:28.640 from Melbourne. 00:00:29.359 >> Agency is so powerful. What's going to 00:00:31.279 happen? Hi, I'm Finnlay and I'm one of 00:00:32.880 the co-founders of Fluency and this is 00:00:34.559 the best podcast ever made. Founders in 00:00:36.559 Motion. [music] 00:00:40.640 >> Quick thing before we get started. We 00:00:42.800 have a huge goal this year of hitting 00:00:44.800 10,000 subscribers. Yes, it's ambitious, 00:00:48.160 but it lets us help more people build 00:00:50.640 really great companies. So if you enjoy 00:00:53.039 the content, learn something new, the 00:00:55.360 best way to support us is by 00:00:56.960 subscribing. Okay, let's get back to So 00:00:59.600 let's start at the most recent news. You 00:01:01.840 just raised 6 million from Accel, but 00:01:04.319 had only been in the US for a few weeks. 00:01:06.560 You weren't even fully booed in yet. So 00:01:09.200 walk me through how the round happened. 00:01:10.960 >> We weren't really sure we were going to 00:01:12.479 raise at the beginning. 00:01:13.920 >> Mhm. we have enough traction to not 00:01:16.240 really like a raise wasn't imperative 00:01:18.240 for us at the time and we're like let's 00:01:20.720 just write two pages send it to some VCs 00:01:23.280 and see what happens that was it and 00:01:25.040 then we sent like a literal two pager we 00:01:27.200 had no deck no data room didn't have a 00:01:29.759 data room till right at the end for DD 00:01:31.600 with Accel I sent it to like 10 funds 00:01:33.759 had some conversations got some intros 00:01:36.000 probably spoke to 18 funds or less total 00:01:38.880 for the whole thing and I'd met like a 00:01:40.720 couple of US funds just in my travels 00:01:43.280 generally 00:01:44.146 >> [snorts] 00:01:44.159 >> when I went to the US with Ollie like 00:01:48.000 7 months ago. 00:01:48.799 >> Put it into context. Say if an 00:01:50.159 Australian founder is looking to raise 00:01:52.079 from US investors, what is your 00:01:55.040 recommendation in terms of the best 00:01:56.560 things that helped you guys? 00:01:58.560 >> First of all, two misinterpretations 00:02:00.399 from Australians raising capital in the 00:02:02.159 US. 00:02:02.640 >> Yes. 00:02:03.360 >> One is the fact that they think it's 00:02:04.880 like really easy. You just go to the 00:02:06.479 valley and you can just raise a massive 00:02:07.759 round. 00:02:08.160 >> Mhm. 00:02:09.119 >> I don't think that's inherently true. I 00:02:11.038 met a bunch of founders NSF who like 00:02:12.959 come from other countries like I'm going 00:02:15.040 to raise around and you see them a week 00:02:16.480 later and like I have to go home my 00:02:17.680 visa's up I didn't raise around. So I 00:02:20.080 don't think it's inherently like this 00:02:21.200 magic button that gets pressed. The fact 00:02:23.120 that it's possible like or you can at 00:02:25.120 the same time so I guess it's kind of 00:02:26.640 like a cognitive dissonant sort of piece 00:02:29.360 like there's everyone kind of thinks 00:02:30.720 like really easy but then like you can 00:02:32.480 just go and do it as well. They kind of 00:02:34.080 both coexist together. What worked for 00:02:35.760 us though is like one I was like I am 00:02:37.200 moving I've moved here. 00:02:38.560 >> Yeah. whilst I was in the process of 00:02:40.400 literally moving there like regardless 00:02:41.760 of whether we raised or not. 00:02:43.680 >> The second was the attitude we had 00:02:46.000 politely was we have enough capital in 00:02:48.480 the bank anyway and we're going to get 00:02:50.160 it done. Thirdly, it was that we did 00:02:53.440 have a good amount of traction on the 00:02:55.280 new product. 00:02:56.160 >> And then the fourth piece was for Accel 00:02:57.840 specifically, 00:02:59.440 >> what we were building perfectly hit a 00:03:01.120 thesis they had internally. 00:03:02.400 >> Yeah. 00:03:02.640 >> So that was some serendipity. So, it was 00:03:04.640 like meeting one with the partner and 00:03:06.959 then like 2 hours later. Maybe that's a 00:03:08.720 bit of hyperbole. It felt like 2 hours 00:03:10.239 later within a 24-hour period. 00:03:12.080 >> Mhm. 00:03:12.560 >> There was like another separate partner 00:03:14.239 meeting. 00:03:14.800 >> Yeah. 00:03:15.440 >> And then it was like I see the next day 00:03:17.360 with like 24 partners on the call. 00:03:19.440 >> Yeah. 00:03:19.840 >> And then it was like a lot of back and 00:03:22.319 forth between myself and our investor, 00:03:23.760 our partner. And they were awesome 00:03:25.120 actually. They were 00:03:28.000 >> so sophisticated in their relationship 00:03:30.799 building. I would say like they were 00:03:32.799 conscious that there was a deadline 00:03:34.400 approaching 00:03:35.760 >> and they were really trying hard to 00:03:37.440 build a genuine relationship despite the 00:03:40.400 timeline. 00:03:41.200 >> Yeah. 00:03:41.840 >> Um and then we [clears throat] went to 00:03:43.040 DD which was 12 background calls pretty 00:03:46.480 much just on myself. They called my 00:03:48.000 brother whilst he was so my brother's a 00:03:49.680 chippy. They called him whilst he was on 00:03:51.440 site. There like hammers and stuff going 00:03:53.360 in the background. Ex employers, 00:03:55.280 customers, friends, other founders. So 00:03:58.480 like yeah like six odd hours of 00:04:00.080 background calls data room DD but 00:04:02.480 there's not much in the data room so 00:04:03.760 that was pretty chill and then they were 00:04:05.360 like helping us find the right other 00:04:06.879 partners to help to to jump in. Biggest 00:04:09.040 thing we wanted was like talent 00:04:10.959 attractiveness because we're 00:04:12.560 >> unknown people from relatively unknown 00:04:14.879 unis especially over in the US. 00:04:17.040 >> So we had no network over there to 00:04:19.040 attract really great talent. So yeah. 00:04:21.120 >> So okay you've already had Australian 00:04:23.120 customers like A PBH and Spec Savers on 00:04:26.639 board. Um, there is this notion going 00:04:29.840 around about US investors discounting 00:04:32.479 Australian revenues. So, did you ever 00:04:34.800 feel that US investors discounted your 00:04:37.440 revenues because they were in Australia? 00:04:39.680 >> No, not for us. 00:04:41.120 >> Yeah, 00:04:41.759 >> cuz I we'd heard that a lot as well and 00:04:43.280 I was like, "Oh, they're going to 00:04:44.160 discount it to zero." Um, maybe if you 00:04:47.040 like vertical software, 00:04:48.880 >> but we're like extremely horizontal. 00:04:51.120 Like one of our Achilles heels is like 00:04:53.199 how big the opportunity space is for 00:04:54.720 what we're building can do. 00:04:56.000 >> Yeah. in terms of like you have to make 00:04:57.120 the right part decisions blah blah blah 00:04:58.240 but back to answering the question is no 00:05:00.320 I I don't think we really had a massive 00:05:02.000 amount of discounting to zero 00:05:04.080 >> to be honest um Accel didn't anyway the 00:05:07.199 revenues maybe half and half 00:05:09.680 >> at the time of the seed but the big 00:05:11.360 logos were the US side 00:05:13.840 >> maybe even less than half and half no 00:05:15.039 it's probably more predominantly 00:05:16.080 Australian revenue split but the bigger 00:05:18.479 logos were on the US side 00:05:20.160 >> so now that you've experienced both 00:05:21.759 countries Australia and the US what do 00:05:24.000 you think is the hardest thing about 00:05:26.000 building in Australia the founders face. 00:05:28.720 >> I've just got back from the valley a 00:05:30.960 couple weeks ago. First time back to 00:05:32.240 Australia in like 5 months is I just 00:05:34.000 want to put my feet up and kick back. 00:05:36.080 Like there's like no 00:05:38.560 exttrinsic pressure here. Like in the 00:05:40.479 valley it's like if you're not working 00:05:41.520 you feel like you're falling behind and 00:05:43.199 that's awesome. Like that's what you 00:05:44.639 should optimize for if you are a founder 00:05:46.320 and you do want to have like a venture 00:05:47.759 scale outcome. 00:05:48.720 >> I think it's a positive thing. 00:05:49.759 >> Okay. So maybe let's zoom out for a 00:05:51.360 second. What does fluency actually do? 00:05:54.400 How it works is Fluency is deployed onto 00:05:57.440 let's say everybody's laptop inside of 00:05:59.759 an organization and for all intents and 00:06:02.560 purposes 00:06:04.000 >> we see everything that happens. We take 00:06:06.080 screenshots. We scrape like system logs. 00:06:08.000 We get ingress egress for like network 00:06:10.160 logs. We get a lot of information 00:06:13.440 and we're getting this every day as time 00:06:15.759 passes across the whole organization. We 00:06:17.840 build what's called a work ontology. 00:06:20.240 >> Mhm. 00:06:21.600 An ontology is just like a tech word for 00:06:24.960 a queryable graph where a graph is a 00:06:29.360 form of data store where you have nodes 00:06:32.080 and edges. A node represents some data 00:06:34.800 and an edge is the relationship. So this 00:06:36.800 is being built over time and it's 00:06:38.080 constantly changing depending on new 00:06:39.600 context we get. It's non- monotonic, 00:06:42.080 which means that the the present can 00:06:44.400 affect the past because sometimes you 00:06:45.919 might get an email or a communication or 00:06:47.680 something changes in your business which 00:06:49.600 affects what we thought was the truth 00:06:52.240 previously. 00:06:53.600 >> And there's three levels to our 00:06:55.520 ontology. The first one defines what 00:06:58.160 work is. So we're like working on 00:07:01.039 software or this is a process for 00:07:02.880 invoicing. Just making sure we know 00:07:04.639 who's who in the zoo. like cuz like to a 00:07:07.360 computer perspective you are different 00:07:09.599 in emails in Slack in the CRM that you 00:07:12.560 use but we just want to make sure that 00:07:13.919 we always know it's you 00:07:15.440 >> and the third layer is we understand 00:07:18.240 what is being built through these 00:07:19.840 processes you know whether it's an 00:07:21.199 invoice or a PowerPoint presentation or 00:07:24.319 a unit of code what that empowers from 00:07:26.400 an outcome perspective today is a couple 00:07:29.759 of things so there's like three main use 00:07:31.280 cases for fluency 00:07:32.720 >> the first cab off the rank is businesses 00:07:34.880 don't actually know what their process 00:07:35.759 processes are how they get done, what 00:07:37.199 are the edge cases, what are the 00:07:38.479 exceptions, if you know this weird tiny 00:07:41.520 bug happens, I have to elevate it over 00:07:43.599 here, whatever. We capture and give them 00:07:45.680 that information. 00:07:47.120 >> The next value prop is once we've found 00:07:50.400 all of your workflows, processes, and 00:07:52.400 tasks, we tell you how to improve them. 00:07:54.479 Usually that's through automation. So we 00:07:57.520 give them the map, the data, the APIs to 00:08:01.280 build that automation on top of our 00:08:02.960 platform in you know a very short amount 00:08:05.680 of time accurately. Whether it's aentic 00:08:08.000 or otherwise doesn't really matter. 00:08:09.840 Flavor of the day is agentic though. 00:08:11.520 >> Yeah. 00:08:11.919 >> The final use case we enable today is 00:08:14.800 measuring the ROI on change itself. 00:08:17.199 >> Yeah. 00:08:17.759 >> So this to put it in a question. I'm a 00:08:20.720 CFO at Acme Core and I want to answer 00:08:23.599 the question for people in my team who 00:08:26.560 are the biggest power users of Microsoft 00:08:28.560 Copilot. How much more effectively do 00:08:30.720 they get their work done? How much 00:08:32.399 capacity is freed up? Do they access 00:08:34.000 documentation less? We can answer any of 00:08:35.839 those sort of questions and then you can 00:08:37.679 attribute it to financial dollars or 00:08:39.200 generic productivity tracking in 00:08:40.719 aggregate because we don't allow 00:08:41.919 individual employee monitoring or 00:08:44.000 productivity indexing. Um but in layman 00:08:46.880 terms what fluency does is they look at 00:08:50.399 how work is done across your 00:08:51.920 organization. Figure out what are your 00:08:54.399 processes where the areas that you can 00:08:56.640 optimize further and create a more 00:08:59.360 intelligent workflow for all 00:09:00.800 organizations. 00:09:01.680 >> Couldn't have said it better myself. 00:09:03.040 That's what I meant. [laughter] 00:09:05.360 >> You can steal that next time. 00:09:06.560 >> Yeah. Yeah. I will. And that's just 00:09:08.160 today. 00:09:08.800 >> And that's just today. So this is 00:09:11.040 already like a really huge product and a 00:09:13.200 really huge vision to understand how 00:09:15.120 work is being done and how to make it 00:09:17.120 better. Do you see an even further 00:09:19.680 evolution of fluency? 00:09:21.200 >> Yes, absolutely. We kind of expand the 00:09:24.000 vision every month. What it is today and 00:09:25.519 what we've started working on is 00:09:28.640 actually step 1.5 is we will run the 00:09:30.880 automations ourselves on top of our 00:09:32.320 ontology. So the ontology or the map of 00:09:35.279 work is really really good at helping 00:09:37.600 you automate things without needing 00:09:39.040 developers, without having to worry 00:09:40.560 about agents hallucinating as much 00:09:42.959 because it has all of the examples of 00:09:44.720 all that ever was. But it still requires 00:09:47.279 a good amount of setup and work to do 00:09:48.800 that. It's not entirely accurate. 00:09:51.760 What we're moving towards next is 00:09:53.279 building our own foundational models. 00:09:55.839 Not large language models though, what 00:09:57.360 are called world models. So a large 00:09:59.360 language model, as I'm sure everyone's 00:10:01.040 used chat GPT, it predicts the next 00:10:04.160 sentence, word, what's called a token, a 00:10:06.160 small subset of text. It's effectively 00:10:08.720 predicting like the future in a very 00:10:10.959 discreet discrete manner. What a world 00:10:14.079 model does, it uses the same 00:10:15.600 architecture to predict the next most 00:10:18.560 likely event. 00:10:20.480 >> It's still trained on text or like 00:10:22.640 binary because that's what a computer 00:10:24.000 is, but it's not representing text. 00:10:25.360 It'll represent something like a video 00:10:26.880 game or represent a manufacturing floor. 00:10:30.000 For us, it represents how work gets done 00:10:32.560 in a business. So, it's graph based. 00:10:34.399 It's that work map based. To rationalize 00:10:37.440 that, you put work map in and it'll 00:10:41.279 predict what's going to happen next. 00:10:42.640 Work map out. So, what we think is going 00:10:45.920 to happen inside of our early R&D is you 00:10:48.720 can have like effective super agents on 00:10:50.399 top of your process. It can run it all 00:10:52.399 like fully autonomously almost zeroot 00:10:55.760 workflows and orchestrated workflows 00:10:58.720 like almost like um department based 00:11:02.000 work without needing to like set it all 00:11:04.079 up. And the other piece is you should be 00:11:05.440 able to put in like what a future state 00:11:07.519 looks like into our world model 00:11:09.600 >> and we'll predict what's going to happen 00:11:11.440 like how is it going to affect the team? 00:11:12.640 How's it going to affect your revenue? 00:11:13.680 How's it going to affect your finances 00:11:14.800 productivity? All of that sort of piece. 00:11:16.800 >> Yeah. 00:11:17.120 >> So that's where we're heading and we 00:11:18.399 believe the best way to do this is like 00:11:19.839 a sovereign 00:11:21.279 world model per customer. So most of our 00:11:23.600 customers are more larger enterprises 00:11:25.200 which would be able to afford this. 00:11:26.880 >> Yes. So my simple understanding here is 00:11:30.320 that because you are able to build such 00:11:32.480 a deep context map of how work gets done 00:11:34.480 and how operations are currently done, 00:11:36.800 you have kind of you have a very unique 00:11:40.000 understanding of the organization. Then 00:11:42.160 you're able to predict better because 00:11:45.839 it's [clears throat] trained all on that 00:11:47.120 kind of unique data set. 00:11:48.399 >> Yeah. like you send an email and we see 00:11:50.399 in a time series graph basically what 00:11:52.959 happens next every single time and that 00:11:55.200 happens for anything you do on your 00:11:57.279 machine we abstract it and build it out 00:11:59.360 for the whole organization one of our 00:12:01.120 investors heads up AI at Princeton and 00:12:03.440 she's super focused on building out 00:12:05.040 world models so working closely with her 00:12:07.120 to you know make sure we do it right 00:12:09.519 >> yeah yeah and like when you think about 00:12:11.360 this next stage of fluency the 1.5 or 00:12:14.160 the number two act when do you think 00:12:17.440 this can see the light of day 00:12:19.279 >> 1.5 8 to 12 weeks. 00:12:21.279 >> Mhm. 00:12:21.920 >> Two 9 months. 00:12:23.680 >> Speaking of like uh agents and your idea 00:12:26.399 of like super agents, I remember a while 00:12:28.160 back you said to me that agentic AI is 00:12:31.120 currently just robotic process 00:12:32.800 automation on the cloud and that VCs are 00:12:36.480 overhyping it and getting it wrong. So 00:12:38.880 this is kind of a hot take. Do you still 00:12:41.839 agree with this and why do you think the 00:12:43.839 current hype cycle is missing the mark? 00:12:45.760 >> I don't agree with it anymore. 00:12:47.013 [laughter] 00:12:48.639 Yeah, I [snorts] don't agree with that 00:12:49.680 anymore. I think the opposite's now 00:12:51.040 true. 00:12:51.360 >> Yeah, 00:12:51.760 >> I think pretty much everybody is quite 00:12:53.360 flippant about the current capabilities 00:12:54.880 and like like when I said that I think 00:12:57.200 it was probably true at the time. 00:12:58.720 >> Yeah. 00:12:59.680 >> And now I completely disagree. I think 00:13:02.079 we've gone the other way. I think we're 00:13:04.000 like undervaluing especially as like a 00:13:06.240 broader people outside of tech. 00:13:08.079 >> Yeah. 00:13:08.959 >> What is coming 00:13:10.560 >> Yeah. 00:13:10.959 >> what's possible. Like assuming there's 00:13:14.320 no like mass extinction event in our 00:13:16.079 lifetimes. Mhm. 00:13:18.160 >> There's probably multiple futures where 00:13:19.839 you could say like the only limit is 00:13:21.440 really our imagination. 00:13:22.800 >> Yeah. 00:13:23.360 >> It first start in like computer space 00:13:25.839 where the only limit is your 00:13:26.800 imagination. And then once we have these 00:13:30.000 world well actually world models but for 00:13:31.680 physical reality where they're like 00:13:32.800 augmenting a humanoid robot or I don't 00:13:35.120 know some factory thing. 00:13:36.959 >> Well, what do you think the workforce 00:13:38.560 will be in the next 10 years? 00:13:39.920 >> I agree. Roles will be created. 00:13:41.600 >> Mhm. 00:13:42.320 >> But I don't think it'll match the 00:13:43.519 displacement that'll happen. 00:13:44.959 >> Yeah. as in like I don't think Oh, maybe 00:13:47.360 though, as I said, you know, the 00:13:49.680 abstraction floor comes up. 00:13:51.120 >> Yeah. 00:13:51.440 >> The abstraction ceiling comes up with 00:13:53.839 it. Maybe when we get to that point, 00:13:56.000 there will be sufficient rolls created 00:13:59.360 to supplement the displacement that'll 00:14:01.120 likely occur. 00:14:02.320 >> Yeah. 00:14:02.800 >> Um I can't see it right now though, but 00:14:05.120 that's probably because like the 00:14:06.160 acceleration is faster than the floor, 00:14:08.399 and there'll be like some sort of 00:14:09.519 uncomfortable period, but then it'll 00:14:10.880 come up to it. But maybe we'll just 00:14:12.800 become like really really productive. 00:14:14.880 Mhm. 00:14:15.519 >> And we'll learn to like work with this 00:14:18.000 much productivity. So they'll just be 00:14:19.279 like 00:14:20.079 >> same amount of jobs. 00:14:21.279 >> Yeah. 00:14:21.519 >> But everyone's just like 100x more 00:14:22.959 productive. 00:14:23.600 >> Okay. So you've had your fair share of 00:14:25.440 pivots. Before the current iteration of 00:14:27.600 Fluency, you were doing standard 00:14:29.440 operating procedure capturing. 00:14:31.920 >> What broke that made you realize that 00:14:34.079 the first version wasn't going to work? 00:14:35.839 We had the plan of going from what we 00:14:37.760 called pickum process creation module 00:14:39.440 which was SOP creation to POM process 00:14:42.160 observation module which is kind of 00:14:43.519 where we were today to PAM which was 00:14:45.360 process automation module which is also 00:14:47.440 kind of where we are today. SOP product 00:14:50.399 yeah got a decent amount of revenue got 00:14:52.160 some decent customers but it wasn't we 00:14:54.000 didn't feel the product market fit. 00:14:56.160 Didn't even feel like we really had 00:14:57.199 inklings of product market fit. 00:14:58.880 Everything felt like grind. 00:15:00.639 >> Yeah. 00:15:01.040 >> Like for selling that product it was 00:15:02.639 just like Copus pushing the ball up the 00:15:04.720 hill. for every sale [snorts] 00:15:07.519 >> there was a use but the buyer like the 00:15:10.240 user wasn't really the buyer and the 00:15:12.240 buyer didn't have that much pain 00:15:14.320 associated with the user there were some 00:15:16.240 companies who were really appreciative 00:15:17.600 but they were few and far between 00:15:19.839 >> um so we're like let's just bring 00:15:21.440 forward and see how that goes 00:15:23.360 >> and then [clears throat] we slowly 00:15:24.720 worked on it for a while took a ton 00:15:27.120 of time to work out how to get it to 00:15:28.959 work 00:15:29.519 >> cuz like how do you define what work is 00:15:31.600 subjective 00:15:32.399 >> y 00:15:32.880 >> how do you know what is a process and 00:15:34.079 what isn't a process across anything 00:15:36.399 anyone could possibly be doing on a 00:15:38.079 computer. 00:15:38.639 >> Yeah. 00:15:39.040 >> When a process is really like a human 00:15:41.920 applied definition for a pattern of 00:15:44.480 action. 00:15:45.279 >> Yeah. 00:15:45.760 >> So trying to work out all those problems 00:15:47.360 eventually got it to work. We went to 00:15:49.040 market with it and within one week we're 00:15:50.480 like okay we're hard pivoting because 00:15:52.160 the reception was completely different. 00:15:53.920 It was like all right this is definitely 00:15:56.160 what we should do. M 00:15:57.600 >> so the old product still kind of exists 00:16:00.560 in in the new product today. But it's 00:16:02.079 like if you assume building process 00:16:04.800 documentation was 100% of the old 00:16:06.959 product. That's like 00:16:09.600 10 basis points like.1% of the new 00:16:12.399 product in terms of like capability and 00:16:14.320 value creation. 00:16:15.440 >> So we're still maintaining it. We still 00:16:17.199 still got a couple of deals. Still got a 00:16:19.040 lot of customers on it. Um most of them 00:16:21.360 still love it to be honest. 00:16:22.800 >> Yeah. 00:16:23.199 >> But we're very much focused on Yeah. 00:16:25.360 Fluency. We just called it Fluency 00:16:27.199 again. 00:16:27.519 >> Okay. So, when you first started 00:16:29.120 Fluency, you were in your early 20s 00:16:32.079 trying to sell to really large 00:16:33.759 enterprises in Australia and like you 00:16:35.839 said, like pushing up the hill like 00:16:39.519 >> did anyone ever look at you and just not 00:16:41.680 take you seriously? 00:16:43.519 >> Honestly, I never felt that to be 00:16:45.920 honest. No, not really. 00:16:47.279 >> I never felt like I got disparaged 00:16:49.759 against my age. M 00:16:51.279 >> I never felt we got disparaged against 00:16:54.320 >> unreasonably as a startup. 00:16:56.240 >> Like I always felt like the businesses 00:16:57.759 were de-risking appropriately. Like I 00:16:59.360 never like 00:17:00.800 >> Yeah. Yeah. I mean at the core of it, 00:17:03.040 right? It's solving a problem for a 00:17:05.199 customer 00:17:06.079 >> and it's amazing when that problem is 00:17:08.079 like the number one problem that they're 00:17:09.599 thinking about on a day-to-day basis. 00:17:11.199 >> If the large language model is the 00:17:12.959 ocean. 00:17:13.599 >> Yeah. 00:17:14.559 >> And the waves are the problems. 00:17:17.280 >> We're like surfing one of the waves 00:17:18.720 right now. M 00:17:19.760 >> but hopefully we'll become part of the 00:17:21.119 ocean too. 00:17:22.000 >> Yeah. 00:17:22.240 >> And create waves. 00:17:23.439 >> Yeah, why not? Okay. And then so you 00:17:26.240 went from Melbourne to SF. In Australia 00:17:28.559 there's this thing called toe poppy 00:17:30.160 syndrome. 00:17:31.200 >> Um it's essentially when people hate 00:17:33.360 when you stand out or are too ambitious. 00:17:35.679 I think you actually told me about this. 00:17:37.600 I didn't even know it existed. 00:17:39.440 >> Did you feel that living here? And how 00:17:41.760 is that different in ASF? 00:17:44.400 >> I actually don't think I or we 00:17:48.559 >> as fluency has experienced that much 00:17:50.240 here in Australia. The only like times 00:17:52.240 I've seen tall poppy syndrome in 00:17:53.840 Australia 00:17:54.640 >> is either when someone obviously has 00:17:56.559 money and has like a really nice car. 00:17:58.240 Like I've got a mate who's 22, really 00:18:00.000 successful, just bought like a half a 00:18:02.000 million dollar Porsche. 00:18:03.039 >> Mhm. 00:18:03.440 >> And people like old people like scoff. 00:18:05.679 You can see them scoff when you're in 00:18:06.720 the car [laughter] 00:18:08.000 and they're like this like assuming it's 00:18:09.520 parents money. I think that's actually a 00:18:11.039 form of tall puppy syndrome [snorts] 00:18:12.720 like you don't assume that the person 00:18:14.799 like worked hard for it etc. Um the 00:18:17.760 other one is more like I don't know if 00:18:19.840 this is tall poppy syndrome. I think 00:18:21.919 it's just more like Australian lifestyle 00:18:23.520 propagation but it's like 00:18:25.440 >> when you make work your priority 00:18:29.120 >> and do 14 hour days you know six days a 00:18:31.919 week and you don't really take much time 00:18:33.440 off you hang out with your friends less. 00:18:36.000 There's a bit of like, oh, why the you 00:18:37.679 doing that? Like, it's your 20s. Like, 00:18:38.960 you're wasting your 20s. 00:18:40.400 >> Um, 00:18:41.520 >> there's a little bit of that, but I 00:18:43.520 think it's just like wrong frame of 00:18:44.480 mind. Like, if you're a professional 00:18:45.440 tennis player player in Australia, 00:18:47.120 people like, 00:18:47.919 >> "Yeah, keep it up." Like, don't you dare 00:18:50.000 drink that beer. [laughter] 00:18:51.679 >> But then to go back to the work side, I 00:18:54.320 mean, again, super privileged, 00:18:55.600 obviously, very lucky, but I get to 00:18:56.960 travel like kind of around the world, 00:18:58.480 around the States, new places, meet new 00:19:00.720 people, see all that sort of stuff. So, 00:19:02.960 I don't feel like there's any 20's 00:19:04.640 wasting, but they're like the only three 00:19:06.480 examples I've seen of tall poppy 00:19:08.080 syndrome, and I don't really feel I've 00:19:09.760 experienced much. A little bit of number 00:19:11.440 three, 00:19:12.400 >> but I don't really care. 00:19:14.320 >> Yeah, I agree. I actually don't see it 00:19:17.280 too much during my time. Um, and and 00:19:20.160 yeah, but I do agree with the third 00:19:21.679 point in the sense that like sometimes I 00:19:23.120 get asked like, "Why are you working so 00:19:24.559 hard?" blah blah blah, but it's really 00:19:26.720 actually it's just really fun to have 00:19:28.080 conversation with people that are way 00:19:29.360 better than you and you can learn 00:19:30.480 something new. And that is actually 00:19:32.640 pretty fun and interesting. 00:19:34.640 >> Like we're lucky that I guess like work 00:19:36.160 is one of our hobbies. 00:19:37.919 >> Like it actually 00:19:41.440 makes me money. It makes me feel 00:19:43.679 fulfilled. I get to like have impact. I 00:19:45.840 know the team feels that way. Like we 00:19:47.840 >> like the opportunity space for which 00:19:49.120 we're solving S could be huge. 00:19:51.360 >> That's exciting. It's like why wouldn't 00:19:52.720 we grind? 00:19:53.840 >> Yeah. You know, like 00:19:54.720 >> still exercise, still friends. 00:19:57.120 >> Yeah. When you're tired just like invite 00:19:58.480 them over to the speak easy, right? 00:20:00.240 >> Exactly. The fluency bar, the flaw. 00:20:02.400 >> Yeah, that that that's like the hot 00:20:04.000 place to be in SF right now. 00:20:05.520 >> It is SF's most hidden bar. 00:20:07.600 >> Okay. So, you raised 6 million. That 00:20:10.080 means people are expecting a unicorn or 00:20:12.720 decagon outcome. 00:20:14.880 >> Do you ever feel the weight of that? 00:20:19.200 >> Um, no. I feel like investor pressure a 00:20:23.120 little bit, 00:20:23.679 >> but I'd actually think that's more like 00:20:25.039 internally manufactured. All of our 00:20:26.480 investors are quite chill. 00:20:27.840 >> Mhm. 00:20:28.159 >> They're like, you know, let's work it 00:20:29.120 out. Obviously, work hard, we'll help 00:20:30.480 you where we can. I feel like an 00:20:32.320 invisible investor pressure that I don't 00:20:33.840 think actually really exists. Yeah. 00:20:35.600 >> But I don't feel the weight of 00:20:38.799 >> having to be a billion dollar outcome or 00:20:40.720 $10 billion outcome because 00:20:42.880 >> for us as a team, we're all aligned on 00:20:44.720 the fact that we want to try to get 00:20:46.000 there. 00:20:46.480 >> Yeah. 00:20:46.720 >> And we would rather die trying like 00:20:48.799 company die trying. 00:20:49.840 >> Yeah. 00:20:50.480 >> Than be like a $50 million company. 00:20:52.559 >> Yes. 00:20:52.880 >> So we would rather spend all of our 00:20:54.159 money and die at 100 million. 00:20:55.679 >> Mhm. then, you know, play the more 00:20:58.559 conservative game, nickel and dime, move 00:21:00.400 to a profitable model earlier and be the 00:21:02.400 $50 million$100 million company. 00:21:04.080 >> Yeah, 00:21:04.480 >> we all want that. We've made our peace 00:21:05.760 with that. 00:21:06.480 >> So, I don't really feel 00:21:08.400 >> the pressure cuz I'm kind of aligned 00:21:10.000 there, you know. I want to try and get 00:21:11.520 that done. 00:21:12.320 >> Yeah. 00:21:12.640 >> So, we'll do our best to get there. And 00:21:14.400 if we don't, I don't know. Cost of 00:21:15.919 playing venture. 00:21:16.880 >> Yeah. And then the other thing, too, is 00:21:19.120 like you're 00:21:20.799 >> You will. You will. Of course. Um you're 00:21:23.760 working so hard. You're traveling 00:21:25.600 everywhere at every other week. You've 00:21:29.760 basically uprooted your whole life from 00:21:31.440 Melbourne to SF. Do you ever feel like 00:21:36.159 lonely or missing in your personal life? 00:21:39.120 >> I miss my family. I miss my brother. 00:21:41.039 >> Yeah. 00:21:41.280 >> I miss my family's pets. I miss my 00:21:42.799 girlfriend. 00:21:44.240 >> That's about it. 00:21:46.320 >> I miss my friends as well, actually. No, 00:21:47.919 I miss my friends back here a little 00:21:49.440 bit. 00:21:49.919 >> Yeah. 00:21:50.320 >> But 00:21:51.120 >> I mean have I like the people I work 00:21:53.039 with. M 00:21:54.159 >> a bunch of my friends are my colleagues, 00:21:56.400 >> so it hasn't been so bad. When I first 00:21:59.120 moved, I definitely really missed my 00:22:00.640 like nuclear family, like my parents and 00:22:02.240 my brother, cuz we're all quite close. 00:22:04.000 >> Yeah. 00:22:04.480 >> But kind of made peace with it now. 00:22:06.240 >> Building fluency has cost you something. 00:22:08.720 What do you think that is? 00:22:13.120 >> My skateboarding ability. 00:22:15.200 >> Here are you really good? 00:22:16.799 >> I wasn't. No, 00:22:19.039 no, I was okay. I really enjoyed it. 00:22:21.440 mountain biking ability, I guess, like a 00:22:23.840 more extended friend network. Like 00:22:25.679 friend circles become a lot closer now, 00:22:27.679 but kind of happy with that as well. Um, 00:22:32.159 obviously like time with family, that's 00:22:33.840 probably like the number one cost to be 00:22:35.360 honest. 00:22:35.919 >> If you had to redo this entire process 00:22:38.080 all over again, building fluency from 00:22:40.240 scratch. So like going back to 2023 00:22:43.039 Swinber accelerator days 00:22:45.120 >> with what I know now, 00:22:46.080 >> with what you know now, 00:22:47.520 >> what's like the one thing you would do 00:22:49.440 differently? You can be like super 00:22:51.200 ambitious from day one and you can just 00:22:52.799 do things. I know like everybody says 00:22:54.480 that, but like agency is so powerful to 00:22:57.760 just like do Just like try and do 00:22:59.120 like what's going to happen. 00:23:00.400 >> Hey, I love that. Um, agency, guys. 00:23:02.960 That's a wrap. If you like this episode, 00:23:05.039 please hit the like and subscribe 00:23:06.320 button. It helps us bring on more 00:23:07.919 awesome guests, level our production, 00:23:09.760 and bring on new series you'll want to 00:23:11.520 watch. And if you want to hear more 00:23:13.520 early stage builder stories, check out 00:23:15.360 our other episodes. Okay, see you next 00:23:17.760 time.