Invisible Product Podcast Episode 2: PAST - Where Did We Come From?

Summary

The Invisible Product Podcast, hosted by Lisa Cardinal, delves into the company's product evolution with insights from Seb and Casper. Starting as a virtual assistant service for busy executives, Invisible transformed into an operations-as-a-service platform. The episode highlights the company's ability to scale, pivot to meet client needs, and innovate through the development of the Digital Assembly Line (D2), Flow, and the new app. The podcast also discusses the significant growth in the AI training vertical, emphasizing the importance of a flexible and scalable product foundation that has adapted to complex client requirements without a complete overhaul.

Takeaways

• 🎙️ The Invisible Product Podcast is hosted by Lisa Cardinal, a product manager on the customer experience team, and is designed for team members to understand the product stack and processes.
• 📚 The podcast delves into the history and evolution of Invisible's product, with Seb and Casper sharing insights on the company's journey from its inception to the present.
• 🤖 Initially, Invisible was framed as a virtual assistant company, targeting busy executives and offering a single touchpoint for various tasks through a virtual assistant service.
• 🔧 The company built its first proprietary product, the Digital Assembly Line (D2), in late 2018 to manage the high volume of requests from clients, which was later migrated to in early 2019.
• 🔄 Clients began to 'hack' Invisible's service by requesting recurring tasks, which led to the evolution of the product into an operations as a service model, with clients willing to pay more for valuable processes.
• 📈 The demand for on-demand delivery services during the COVID-19 pandemic highlighted the limitations of D2 and led to the development of Flow, a prototype application to manage large-scale operations with multiple agents.
• 🚀 Flow was successful in scaling operations for on-demand delivery clients but was not self-served and required significant engineering effort for onboarding new processes.
• 🛠️ The development of the new app, which combined the scalability of D2 and the user experience of Flow, was a response to the need for a more flexible and scalable system for process onboarding.
• 📊 The pivot to AI training as a significant part of Invisible's revenue came after successful collaborations with OpenAI, which recognized the high-quality data and process management Invisible could provide.
• 🔗 Invisible's product has evolved to focus on configurable interfaces and data transfer capabilities to meet the needs of AI training clients, emphasizing the importance of data quality and process flexibility.
• 🏗️ The foundations of Invisible's product have remained robust, allowing for iterative improvements and scalability without the need for a complete rewrite, which is common in technology companies.

Q & A

What is the purpose of the Invisible Product Podcast?

The Invisible Product Podcast is designed for team members at Invisible, aiming to deep dive into the Invisible product stack and answer questions about the company's products and processes.

Who are the hosts of the second episode of the podcast?

The second episode of the podcast is hosted by Lisa Cardinal, with Seb and Casper as the main speakers who discuss the history and evolution of Invisible's product.

What was Invisible's initial product offering when Seb joined the company in 2018?

When Seb joined in 2018, Invisible positioned itself as a virtual assistant company, offering a single touchpoint for busy executives to delegate tasks via email to a virtual assistant.

What is the term 'delegation' in the context of Invisible's early product?

In the context of Invisible's early product, 'delegation' refers to the tasks or requests sent by clients to their virtual assistant, which Invisible would then parse, work on, and deliver results back to the client.

What was the first proprietary product developed by Invisible and when was it introduced?

The first proprietary product developed by Invisible was the Digital Assembly Line, known as D2, which was introduced in late 2018 and migrated to in early 2019.

How did Invisible's clients start utilizing the service in a way that was not initially intended?

Invisible's clients began to 'hack' the service by requesting recurring tasks and scaling them up, effectively using Invisible as an operations team rather than just for one-off tasks.

What significant event in 2020 impacted Invisible's product development?

The on-demand delivery vertical exploded during the COVID-19 pandemic, which led to a spike in demand for Invisible's services, particularly from companies like DoorDash and GrubHub.

What product did Invisible develop to handle the scale of operations needed during the COVID-19 pandemic?

Invisible developed a prototype application called Flow to help scale a single process and manage up to a hundred agents on that process.

What was the challenge with the Flow application in terms of onboarding new processes?

The challenge with the Flow application was that it was not self-served and required several engineers and weeks to onboard a single process, as each Flow app was bespoke and dedicated to a single process.

What is the significance of the transition from Flow to the new app in Invisible's product evolution?

The transition signifies Invisible's move towards a more scalable and flexible product that can onboard new processes quickly without the need for extensive engineering resources, aligning with the company's direction of operations as a service.

How did Invisible's involvement in AI training impact the company's product direction?

Invisible's involvement in AI training, particularly with clients like OpenAI, led to a significant increase in revenue and a pivot in the company's focus towards supporting complex, flexible, and repetitive client process needs with a scalable and easy-to-use tool.

What feature of Invisible's product has been particularly well-received by clients and the industry?

The flexible interface and the ability to quickly configure and update processes have been particularly well-received, with clients noting the product's ease of use and the 'magic' of real-time process updates.

What is Invisible's strategy for maintaining its product's competitive edge as it scales?

Invisible's strategy includes continuous improvement of its product, focusing on making the tooling as easy to use and scalable as possible, while maintaining a solid and responsive core technology offering.

Outlines

00:00

🎙️ Introduction to the Invisible Product Podcast

The Invisible Product Podcast, hosted by Lisa Cardinal, aims to explore the company's product stack and address team members' queries. The podcast invites questions and topics from the audience via a Slack channel. The second episode features Seb and Casper, who discuss the company's product evolution. Seb, having been with the company since 2018, provides historical context, while Casper, a newer addition, contributes fresh insights. The episode delves into the company's transition from a virtual assistant service to a more complex operations platform.

05:01

🛠️ Evolution from Virtual Assistant to Operations as a Service

Seb recounts the company's journey from a virtual assistant startup in 2018 to its current form. Initially, the company targeted busy executives with a single touchpoint virtual assistant service. As clients began to use the service for recurring tasks, the company adapted by modifying their data model to accommodate recurring and scaled processes. This shift led to the development of the 'Digital Assembly Line' (D2) in 2018 and its migration to a more scalable product in 2019. The company's pivot to operations as a service was solidified by the demand surge during the COVID-19 pandemic, particularly in the on-demand delivery sector.

10:01

🚀 Scaling Challenges and the Development of Flow

The company faced scaling challenges as clients demanded larger volumes of service. This led to the creation of Flow, a prototype application designed to manage a hundred agents on a single process. Flow was instrumental during the pandemic, helping on-demand delivery companies scale their operations. However, Flow's bespoke nature made it challenging to onboard new processes quickly. This realization prompted the development of a new application that combined the user experience of Flow with the scalability of Dal, resulting in a more flexible and efficient system.

15:03

🔄 Pivot to AI Training and the Growth of Invisible

Casper discusses the company's pivot to AI training, which began with a project for OpenAI. The company's ability to provide high-quality human data for training AI models led to significant revenue growth. Invisible's product had to adapt to meet the needs of AI companies, focusing on creating flexible interfaces and efficient data transfer capabilities. The foundations laid by previous developments allowed for these adaptations without a complete overhaul of the product.

20:04

🌟 The Power of Flexibility and the Future of Invisible

The final paragraph highlights the importance of flexibility in Invisible's product, which has become a significant selling point. The company's ability to quickly adapt and update processes has impressed clients, leading to increased demand. The podcast wraps up by emphasizing the product's scalability and the solid foundation that supports it. The team is excited about the product's evolution and the upcoming discussion on the company's current state and workplace story in the next episode.

Keywords

💡Product Manager

A product manager is a professional responsible for the vision, strategy, and roadmap of a product. In the video, Lisa Cardinal introduces herself as a product manager on the customer experience team, indicating her role in guiding the development and success of the company's products.

💡Customer Experience

Customer experience refers to the sum of all interactions a customer has with a company or product. The script mentions Lisa being part of the customer experience team, which is focused on ensuring the product meets the needs and expectations of the customers, enhancing their overall satisfaction.

💡Digital Assembly Line

The term 'Digital Assembly Line' in the script refers to the company's first proprietary product, which was designed to handle a high volume of requests by optimizing the process of ingesting, understanding, and routing tasks to the appropriate agents for execution. It symbolizes the evolution from a virtual assistant model to a more scalable and automated service.

💡Delegation

Delegation in the script is used to describe the process where clients send tasks or requests to the virtual assistant, which then gets parsed and worked on by the company. It's a key concept that originated from the company's initial service model, where busy executives could delegate tasks to a virtual assistant for efficient handling.

💡Operations as a Service (OaaS)

Operations as a Service is a business model where operational tasks are outsourced to a third-party service provider. In the script, the company transitioned from a virtual assistant service to an OaaS model as clients began to rely on them for repetitive and scalable operational tasks, indicating a strategic pivot in the company's service offering.

💡Product Evolution

Product evolution refers to the process of a product changing and improving over time in response to customer needs, market trends, or technological advancements. The script discusses how the company's product evolved from a virtual assistant to a service that could handle recurring and scalable tasks, reflecting the company's adaptability and growth.

💡On-Demand Delivery

On-Demand Delivery is a service model where goods or services are delivered to customers in real-time, often through digital platforms. The script mentions a significant increase in demand for on-demand delivery services during the COVID-19 pandemic, which led to a surge in the company's operations and the need for a more scalable product.

💡AI Training

AI Training in the context of the script refers to the process of providing human-labeled data to train machine learning models. The company found a niche in providing high-quality, human-annotated data for AI training needs, which became a significant driver for their growth and product development.

💡Process Builder

A Process Builder is a tool that allows users to create and modify business processes without the need for coding. The script discusses the development of a Process Builder as a key feature in the company's product, enabling quick and flexible configuration of complex processes to meet diverse client needs.

💡Scalability

Scalability is the ability of a system, network, or process to handle a growing amount of work. The script repeatedly emphasizes the importance of scalability in the company's product, as it allows them to manage an increasing number of requests and processes efficiently, which is crucial for business growth and client satisfaction.

💡Tech Offering

Tech Offering refers to the technological products or services that a company provides to its customers. In the script, the company's tech offering is highlighted as a flexible tool that is superior to others in the market, attracting clients who require a scalable and responsive solution for their operational needs.

Highlights

Introduction to the Invisible Product Podcast with Lisa Cardinal, discussing the company's product stack and evolution.

Seb and Casper provide a historical overview of Invisible's product development from its inception.

Invisible initially positioned itself as a virtual assistant company targeting busy executives for productivity scaling.

The concept of 'delegation' originated from Invisible's service model, where clients could delegate tasks to a virtual assistant.

Invisible built its first proprietary product, the Digital Assembly Line (D2), to manage high request volumes from clients.

Clients began to 'hack' Invisible's service by requesting recurring tasks, leading to the evolution of operations as a service.

Invisible's product had to adapt to handle the scaling of processes for clients, which was not initially designed for.

The development of 'Flow' was a response to the inability of D2 to scale to meet the demands of the on-demand delivery vertical during the COVID-19 pandemic.

Flow allowed Invisible to manage large agent teams on a single process, but had limitations in process onboarding scalability.

The combination of Flow's scalability and D2's ingestion capabilities led to the creation of a new product, 'AI Mantra', in 2021.

Invisible's entry into the AI training vertical was marked by a significant growth in revenue after partnering with OpenAI.

The need for high-quality data and flexible interfaces led to a pivot in Invisible's product direction to better serve the AI training market.

Invisible's product has evolved without needing a complete rewrite, maintaining a sturdy foundation that supports scaling.

The importance of a flexible tool that can quickly adapt to client needs was highlighted as a key selling point for Invisible.

Invisible's ability to quickly modify processes and interfaces in response to client feedback was a significant advantage.

The podcast will continue in the next episode to discuss Invisible's current state and the workplace story.

Transcripts

welcome to the invisible product podcast

my name is Lisa Cardinal and I'm a

product manager on the customer

experience team here at invisible these

episodes are designed for you invisible

team members we're going to Deep dive

into the invisible product stack and we

seek to answer all your questions about

invisible products and processes if you

have any questions or topics that you

want to see us dive into in the future

reach out on the slack Channel called

Product podcast let's dive

in hello everyone Welcome to our second

episode of the product podcast today we

are going to be talking about the past

of our product uh Seb and Casper are

going to take us through how our product

actually came to be and how it's evolved

over the past few years so I'm going to

actually start by just handing getting

it over to Seb since he's been here and

has all that context I yeah you say Seb

and Casper what it's going to be is

seb's going to talk for a long time

because he's been here for a billion

years then I might pop up the last five

minutes and like hi I'm here too but

it's gonna be mostly sad because he's

got all the context that works Casper's

a new kid on the Block exactly um yeah

I'm I'm like take that and you're like

and sync kind of I love the fact that

you like n syn you're new

band yeah I'm giving my age um cool all

right let's talk about it so um yeah uh

when I joined the

company over five years ago now in

August

2018 and um and the the company and the

and the product as a result have changed

a lot in those last five years and I

kind of want to wanted to give everyone

a little bit of insight as to how that

happened because I think it's quite an

interesting story um when I joined the

company if if you went to the website in

August 2018 where I went um when I was

trying to learn about invisible um you

would see all sorts of stuff about

virtual assistants we framed ourselves

very much as a virtual assistant company

um and you know single touch point and

we were targeting um kind of busy execs

who were trying to scale their own

productivity and didn't want to hire you

know 20 interns um and so we gave them

like a single touch point a virtual

assistant um that they could name uh any

anything they wanted Superman or

Spider-Man or you know James Bond or

whatever it was and they could interact

with that assistant over email and they

could just kind of delegate as much

stuff as they want to that assistant um

and then you know invisible would take

those

delegations

um understand them parse them work on

them and deliver results back to the

back to the client um and and these were

this is where kind the word delegation

came from and it was in in that period

of the company was a kind of discret

packet of work that we would kind ingest

do some work on and hand back um very

much in the same style as as a as a

virtual assistant um but and and that

required us to build a product that was

able to that was optimized for ingesting

a lot of requests all the time because

with that kind of service a busy

executive is just sending 10 or 20

emails a day each with a different

request right and so we had to be able

to ingest those understand them parse

them route them and find an agent who

had the skills and the access uh to be

able to execute that and then deliver it

back right and so um we built the Dow um

in late

2018 uh the digital assembly line which

was actually D two um the first digital

assembly line was basically slack uh and

d 2 was our first kind of proprietary um

product um uh and uh and we migrated to

that product in early

2019 um and as I said yeah that was it

optimized to ingest a lot of requests

and those requests were small requests

could usually be done by an agent or two

agents um you know uh and that was fine

so you really didn't need to like create

much structure around um being able to

scale that because it wasn't really

necessary um what you had to scale was

the ingestion of requests right so

that's how d 2 kind of came about and

then what we saw happen over the next

couple years is that our best fit

clients actually started to kind of hack

our service um you know they would come

to us and they would say um hey that

thing that you that you did for me you

know the other the other week can you do

that can you just do that every day from

now on and we would go okay we'll we'll

take that request that delegation and

we'll recur it and we hacked our data

model a little bit and we'd have this

thing just recur and duplicate itself

every day and and then an agent would

pick that up every day and execute it

right then they would come back to us

and they would say hey that thing you're

doing every day can you do it like 10x

now and we go okay I guess we'll have to

you know either that one agent or we'll

have to staff another agent guess and

and these agents now are no longer

taking different requests because we now

have volume on this it was like the same

two agents were now pretty much staffed

to this um

this delegation which was recurring and

scaling and then they would come back to

us and they'd go okay that thing you're

doing can you do it now 100x and at that

point we were like H okay now this kind

of breaks our this breaks our system

it's not what it was designed to do um

but really what they were what they were

doing is they were they were hacking us

right they they' figured out that um

they could use our service to uh as as

as their operations team right and it

that that's how we kind of morphed into

operations as a service uh they tested

us they saw that we were doing good

quality work and they were no longer

giving us kind of menial tasks they were

actually uh using us as operations as a

service and that's kind of how that came

about and and at the same time they were

also willing to pay us a lot more money

to do so um because those processes were

a lot more valuable to them the ROI was

much higher for them

um uh and so that kind of took took a

different uh that that required a

different approach and so um we started

getting more of these kinds of process

clients we also started targeting that

kind of um following that Trend and that

product Market fit um and uh and kind of

modifying our our go to market around

that as well and kind of going after

these you know business processes um and

then in in 2020 that kind of

really blew up with uh with the with the

on demand delivery um uh vertical which

was which absolutely exploded during the

coid

Pandemic those companies um like door

Dash and GrubHub and many of our current

clients their operations teams in

totally collapsed um with the spike in

demand from coid everyone staying home

everyone needing uh needing delivery and

uh and they found an invisible and

Incredibly powerful partner to help them

scale their operations and with that um

obviously the Dow 2 was not able to

handle that kind of scale wasn't built

for it and uh that was when we developed

flow which uh some of you may have heard

of which was a prototype application to

help us scale a single process and be

able to add a 100 agents and manage a

100 agents on a single process and um

and that kind of uh you know that

carried us through the wave of of of ond

demand delivery um uh kind of clients

that that that hit us in in

2020 but we realized that you know um

this was now the direction of the

company and uh you know while Flo while

Flo was accommodating this um what it

really wasn't doing was uh allowing us

to scale that kind of process onboarding

it would take us two three weeks and

several Engineers to onboard a single

process onflow um because every Flow app

was bespoke and for like dedicated for a

single process maybe a couple um but um

but it wasn't self- sered in any way

there was certainly no um there was no

process Builder as as we know today and

so that kind of led us to to to take

that ux of flow and the um the kind of

scalability of Dal and combine them into

into what we now uh what we now know as

the as as our as the as yeah as uh as

mantor as we called it back then and

that's kind of the logic was well if we

can do this but we can onboard it as

quickly as we could on board uh new

delegations in Dow then we have a real

then we have gold um and that is what we

kind of embarked on in 2021 and and

deployed at the end of that year uh and

that's kind of um taken us uh that took

us into this current wave Uh current

wave of of AI training so I joined in

early 2022 which was sort of the the

crossover point between flow and the new

app so we we still we still had lots of

processes on Flow and we were getting

lots of benefit from it but but the

problems as you said Seb it was flow was

highly functional but uniquely

functional because you were you were

hardcoding the interface you were

hardcoding the business flow behind it

in a single use case and therefore if

they needed tweaking then you had an

engineer involved and when you wanted to

get something up that's like hey this is

quite similar but a little bit different

you have to build an entirely new app

and it's just a very very yeah it's

non-scalable in a different way the Flow

app was scalable within a process but it

was not scalable outside a process if a

client suddenly wanted three things it

wouldn't have worked and that's why yeah

and that's one of one of the reasons

actually like when I was joining it was

quite important to me to understand like

what the future held and I remember

having a conversation with Scott about

this and it was like yeah we we've

learned what we're good at and now were

actually enabling us to be good at that

more quickly with more flexibility and

that was that was an exciting piece of

it when I when I was coming

in yeah I and Casper how soon after you

came in did we break into the AI

training vertical yeah so so like I mean

what I think there's an interesting um

phenomenon I think of uh it's I think it

was Gary Player the quote which is the

more I train the luckier I get the

concept that you make your own luck and

I think the history of invisible both as

a company and within the product is

really it's two lucky events that you

also make the most of the first was the

onand delivery it was there was this

huge Spike and indivisible was able to

service door Dash especially and and

handle that and that was the first big

wave of growth and actually the second

piece of luck really kicked off pretty

much the month after I joined it was May

2022 I think it was via Devon's former

housemate long uang who works for um

open Ai and he was like oh yeah I'm

working for this client like you may

have heard of them they're like f away

on mask it's quite cool this company um

and we've got these data trading needs

and oh you do processes a service could

you do this could you provide human data

and and I think it was like well yeah we

could it's a process right yeah sure we

can do that and and we we we talked to

them and I I can't remember who did the

deal I think it was Jay and Cameron with

a k um and uh we ended up doing a bit of

work for them and I think and if you

look at the revenue it goes like May

2022 $5,000 June 2022

$40,000 July and then August 1.8 million

or something stupid like like basically

they what we produced to them was way

better than the data they were getting

from any other client and they were like

yeah we want all of that like give us

everything you can they were just like

yeah feed me um and that and that was

that was the next big growth like and it

was just open initially um and actually

openai were in a situation where they

had they had their own sort of platform

they call it feather now um

may have been called ter back then and

they had a bunch of different interfaces

but what we were providing for them was

um the ability to train and hire and and

execute high quality across a large

number of people which is a gap

especially when you're doing this

complex work but very quickly we

realized hang on like we've got a good

fit here like we need to make sure we

can support this and that's that that

and that's been something that has

really been the sort of it's as a

company it's been a pivot in terms of

that's now a huge amount of our Revenue

but as a the product we we've pushed it

in certain directions in ways that we

maybe weren't thinking about before we

spent a huge amount of time on

interfaces because although open AI we

actually sort of integrate in a loose

fashion with their app so the agents are

working in our application but it's kind

of not particularly tight for the next

ones the next dominoes AWS and coh here

and AI 21 and Microsoft coming up and

another another few that are coming up

like it's all been our own interface and

so that's basically how can we configure

the process within our application so

that we're capturing them data needs uh

we're capturing the ability to we're

handing the ability to do QA at a high

quality and also push this data out and

it's led to a huge amount of like slight

change in Direction on our product we've

spent a lot of effort on these

interfaces and now we're focusing a lot

on how we transfer the data because

those are the primary needs and that's

like that's been what's Driven this

growth and we've got to support it um

but the other key thing is the

foundations were there the foundations

were there for us to actually set up

these processes like the fact that you

have the ability to create a complex

process with a complex interface via

configuration is hugely important like

if you go into the app you'll see just

how many coher processes there are a a21

processes and AWS processes and because

we need the ability to update them to

tweak them very very quickly and also to

handle access to them in a very

controlled manner like the data quality

is hugely important and so all the

pieces that make up the application

actually ironic

apart from the automation piece which

was historically some a large part of

our value proposition um but when you're

doing human data training automation

isn't really a thing like if if you

could automate it they'd be doing it and

they wouldn't be talking to us um so all

the other pieces have have been a huge

part of that and I I I can confidently

say that we would not be able to do this

if we if we didn't have this flexible

interface and this flexible tooling to

man these processes and actually we're

getting to the stage where it's becoming

a huge part of our selling point like

I've demoed it a couple of times this

week and people are saying this is this

is fantastic like this can we use this

like it's moving from the like this is a

tool to support us to this is a huge

part of our offering like this is a huge

part of our Tech offering is the fact

that we have this flexible tool which is

better than other things in the market

right now like scale I know don't have

the flexibility and that's why people

are coming to us they they they've

decided this is what you need and this

is what we'll give you and we are much

more responsive to the fact that this is

a fast moving area where reserch need

change yeah and I remember in the early

days of of um of Amazon as well well

first of all I remember demoing to

Amazon um and demoing the platform to

them right which was actually a platform

that they weren't going to interact with

but they wanted it to see what was going

on behind the scenes and it was it

became clear to them that it was very

powerful and that and that that could

serve their needs and then during

Amazon's onboarding and scaling I

remember um I believe it was was Andre

um who was making changes pretty much on

the daily in Builder um based on

requests that he that we were receiving

from researchers right we were able to

add a data point I remember Persona um

and other uh kinds of sentiment we were

adding data points then be able to

quickly also um modify the UI in a

really flexible way that that uh that

captured those data points and then we

could immediately uh deliver that that

kind of upgraded process and that I

think felt like magic to them because

were not used to that they were used to

um needing to make a request and then

weeks later some code rolled out and and

then they could capture that new data

point and we could literally do it in

the meeting with them um yeah and that

was I think like pretty powerful um uh

from from the from the client's point of

view my favorite story around that and

if you were there you remember was the

the Amazon orchestrations Project where

they had this complex knee that involved

the ability to to basically via

supervisor fine tuning i. a human

providing the raw data mock up

effectively API calls so like something

that would look like the call you would

make to a third party service and then

the return from that and like Victor and

Sange because they're Wizards knocked it

up in about a week and it was a very

stressful week and I wouldn't want to go

back to it but in a week Amazon spent

three months not delivering it and still

we still don't have it we still don't

have their fabled interface and I think

a having fantastic velers helps hugely

but also having a solid framework within

which you can apply these sort of new

interface objects on a process framework

that is solid is hugely valuable like if

if we were trying to build the whole

thing from scratch maybe we'd be three

months in and not having done it but the

fact that actually it's an interface

component but we have the data models

behind it in in our base structure we

have the way of training together

workflows in the Builder canvas we have

all the other pieces so it's it's it's

just one complex bit rather than a full

organism of complex

bits yeah absolutely and I think um uh

you know Builder has gone through a lot

of evolution and kind of in the hands of

pre as well has really

taken made a step change in terms of its

usability and even though that's

something that clients aren't um using

per se um the experience of Builder and

the look and feel of Builder is are now

kind of garnering compliments from you

know people that have been in in the

workflow management business for for a

while um I think I showcased one of

those a couple days ago right from um

what was quote from yeah I've seen a lot

of these workflow systems and yours is

one of the slickest um and yeah I mean

let's not sit here and just be patting

ourselves on the back there's definitely

things we've got to improve and like

we've got a big road map of stuff that

we know is still are still things that

will make people's lives easier like

versioning cross processes to enable

more easy transfer of stuff and we've

got some core infrastructure things that

are going to really make interfaces a

lot more configurable there's there's a

whole lot of stuff that we can do and

it's not just in the Builder space but I

think yeah if if you if you were try and

um sort of from my perspective and

having listened to se as well and heard

these stories before like the first part

of invisible is the story of like

realizing that we are good at servicing

these like

flexible repetitive client process needs

and like trying to then combine the

tooling that we have for like demanding

work in the dial and having the

interfaces and the complexity of

business process in flow into one thing

and the next stage has been how can we

then make the tooling that supports that

as easy to use and as scalable as

possible as we scale so that we have all

these teams who are trying to service

needs but they're independent and they

are running their own mini businesses

almost but enabled by the core

technology

offering yeah excellent I think we're

going to get into that in in the next

episode right Lisa for sure yeah exactly

um one word I wrote down that I kept

hearing you both mentioned over and over

and over again is scaling and the thing

that excited me when I was joining

invisible and talking to team members

was that the product itself has not had

to be fully Rewritten which is like very

common in technology companies is you

um you start with a product you start

meeting

with clients changing it for their needs

and then you get to a certain point

where you're like okay it doesn't

actually meet anyone's needs anymore we

have to start all over again whereas

what I'm hearing with invisible products

is it's very much we're doing that but

we still got that really sturdy

Foundation where it was really well

thought through to support this scaling

as we keep growing we we've done it once

right the the manal project was a FY

right but I think since then yeah the

the the the foundations that were set up

in that the famous paper demo um with

with seb's hands exactly the magic

fingers um that that still holds true

like the the architecture and the the

core like principles behind it I think

have have stood the test of time and I I

feel very good about them continuing to

stand the test of time as we put more

complex but also more refined pieces on

top of

it excellent all right well why don't we

uh wrap here uh next episode we're going

to be talking about the present where we

are today um talking about the workplace

story so if that's something you've been

hearing you've been hearing that word

and you're ready to learn more about uh

stay tuned for the next episode and

thank you both so much for your time and

we will be talking again soon thanks

Lisa thanks

Lisa thank you for tuning into this

episode because these episodes are

designed for you we encourage you to

reach out if you have questions or

topics that you want to see us Di into

in the future use the slack Channel

called Product

podcast special thanks to musician the

re for the use of this

[Music]

song