Building Better Tools for the Writer’s Room with Josh Kline

Hello and welcome. Today we are excited to speak with Josh Kline, who is, in my opinion, a pioneer at the intersection of Hollywood and technology.

‍

Josh is a Emmy-winning technologist who has spent over two decades transforming how film and TV projects get made, from the way directors review daily footage to the way writing teams collaborate on scripts. He co-founded Sample Digital and then developed Digital Dailies, the industry's first cloud system for reviewing shot footage, earning him a 2013 Primetime Engineering Emmy for its impact.

‍

He's since led innovation at companies like Final Draft, Deluxe, and Box, bringing cloud workflows into the creative process. Now, as CEO of Scripto, Josh is reinventing script collaboration with a real-time platform that's powering shows from Stephen Colbert to John Oliver. Josh, thanks for coming on On Production.

‍

Josh Kline (01:19.096)

Cameron, thank you so much for having me. That intro's a lot to live up to. I'll do my best.

‍

Cameron (01:23.99)

I mean, you know, it's been a lot of fun to like dig into the research on you because you have been in the business for a long time and you've been both in technology and in true like Silicon Valley tech startups like Box, not so much a startup anymore, know, publicly traded company that does really great work. But, you know, today I really do want to talk to you about like the backend workflows in scripting and production tools. Cause our audience here.

‍

at On Production cares about the nitty gritty and the utility behind making movies. So I'm curious for you, having witnessed the shift from paper scripts to cloud-based platforms, what do most people not realize about how much complexity goes on behind the scenes in managing a script digitally? Like in other words, how is writing a script in a professional environment

‍

like with a tool like Final Draft or Scripto fundamentally different than just typing in Word. What's happening in the background that a writer or just somebody might not see?

‍

Josh Kline (02:29.006)

Yeah, my background, 30 years ago, I was an independent writer producer. I wrote and sold a script on Final Draft and ended up as the chief strategy officer at Final Draft about, I don't know, dozen years ago. So kind of come full circle. The experience of, you know,

‍

Mostly writers will export a PDF from Final Draft because writers don't like sharing the native file and writers tend to be worried that people will change their words without their authorization. And frankly, that has happened many, many times. Writers take great offense at that. It's disallowed by the guild contracts, but it happens. so writers actually don't share

‍

that final draft files, the final draft being the dominant screenwriting software in the industry. So then the PDF becomes the format for the screenplay. And the problem with that is that the PDF is a publishing format, is not a live data sharing format in any way. You can melt a PDF, you can extract data from a PDF, but it really is meant to be a publishing format. So because

‍

the industry has moved towards connected platforms and data sharing platforms, as you well know for having created one of them. And by the way, I have seen this with RAPID to address shortcomings in legacy platforms that don't share data, that are inefficient in terms of getting data in and out. So what happened is the growth and the creation of a bunch of

‍

other sort of startups that would create tools that would sort of make it so that we're pretending that this PDF is a live format that we can do things with and share with other people because we've got a duct tape and rubber band ecosystem to enable that because it's necessary to create efficiencies in the production process, which is, you know, a hundred plus year old.

‍

Josh Kline (04:55.64)

production system here that we've gone from film to digital. We've moved away from printed scripts to PDFs, but really we've gone from cutting film or even editing on tape to non-linear editing. So all these other platforms have been modernized, right? But where it all starts with the screenplay remains in a legacy state.

‍

So my goal with what I'm doing now is actually to remove that legacy blocker at the head end for entities, for writers, entities being people, production companies, studios, people who care about the data that a script is, right? It's a data type. And that data is more useful. And by the way, I'm fully aware that a writer hearing me say that would cringe. It's a data type.

‍

But it is data, right? The words, the text, that is information that other people...

‍

Cameron (06:00.536)

They're tokens now. That's what people call them now.

‍

Josh Kline (06:05.482)

Somebody just said, by the way, quick tangent, somebody just sent me a link to yet another like AI powered writing company that will write your screenplay, your novel, you just got to buy some tokens. And I'm like, come on. Like, oh yeah, that's what I want. I want to read somebody's AI powered novel. Like this is some of this stuff gets sillier by the day, not more fully baked.

‍

So addressing a script as something that should be dynamic, that should be a data type that is easily shared but secured. So what most people would think is the difference between, know, I'm working in Microsoft Word, I've got a document, and I'm exporting a PDF, because I don't want anybody to mess with it. And the difference would be a generational shift to I'm working in

‍

Google Docs. And it's not a matter of whether you get to see it. It's a matter of whether you can access it to view it, to comment on it, to edit it. It's really about permissions and securing the actual intellectual property that way versus I've baked it into a PDF and that makes me feel better because it's harder to get the data out of it even though anybody on the IT side can melt the PDF in a second.

‍

So that is that's the viewpoint from the position that I was recruited to take, which is a good fit, frankly, because I spent time, as you mentioned, across collaborative systems for filmmakers. I spent time at Final Draft, understand that world quite well. And at Box, being in this massive hyper growth startup and seeing how that machinery works, that is all about content collaboration, cloud content collaboration.

‍

even though it wasn't specific to something like collaborating on scripts. I can assure you I tried when I was inside a box to have us be able to parse that data natively. I just think that this is a part of the media and entertainment production workflow that is in sore need of efficiencies and modernization. it's an exciting thing for me to work on.

‍

Cameron (08:30.88)

Yeah, absolutely. So in software development, we use version control, like Git, to track code changes. In your experience, how does version control for scripts work? And what makes it uniquely challenging in a creative workflow? Like, have you drawn inspiration from software development practices to handle the multiple drafts and revisions of a screenplay? And conversely, are there aspects of script versioning?

‍

like dealing with network approvals or on the fly rewrites that even like coders might find chaotic.

‍

Josh Kline (09:08.462)

I think coders would find it less chaotic. You don't usually have different groups of writers branching off and working on chunks of something and then having to elegantly commit that data, that script or that code back to the corpus. People collaborate.

‍

So different formats are either more or less collaborative. Feature films are usually not written collaboratively. Hour dramas are usually not written collaboratively unless you have a specific writing partner. But as a team, what you get into for collaboration is in the area of sitcoms, very collaborative. And then where scripto comes from, that late night variety space.

‍

is hyper collaborative. that frankly, that was the genesis of Scripto. There wasn't a software solution out there that was collaborative enough for the Colbert report or the Late Show, what became the Late Show for him. And then, you know, all of his buddies liked hearing about something collaborative out there that word, which is where all the early customers came from. So certain data, certain content types, certain

‍

formats are very, very collaborative. Some are less so. In the area of the collaborative ones, and I say across all of the versioning becomes very important. So you reference Git. I mean, that's a great analogy for versioning in script writing. So in our scenario, we take snapshots every few seconds of an activity. We have named drafts as well. And so

‍

We, every draft of every script ever written in Scripto is still in Scripto. And we make it very easy to compare the diff between any two drafts, not just the last one and the current one. Like literally like, I worked on, I did a draft today. I worked on one I think I liked better a month ago. I want to like kind of see where I got off base. How did my okay draft turn into this thing I can't even read anymore?

‍

Josh Kline (11:31.178)

So it is, it's an efficient way to maintain versioning. It's an efficient way to provide a blanket of security for writers. My office at Final Draft was right next door to the head of technical support. And he had a bullpen of tech support personnel right in front of our offices. And they spent a lot of their time.

‍

fixing broken files because third party solutions would open up those files, modify them somehow, and then they work on them with some other tool. And then the writer couldn't open it anymore. It's kind of mayhem. And it speaks to sort of a generational shift on how we deal with sharing data. Like you shouldn't email a file around. That's not generally happening in 2025 in other areas.

‍

either within production or across people who use computers. We're usually sharing information and providing access to things that are working and we're collaborating, but we're not emailing a file to somebody hoping that they'll work on it and send it back to us. Usually, mean, in pockets of activity, but generally speaking, moving away from

‍

Cameron (12:48.952)

Well with that, mean, could you walk me through what the approval process for a script or daily content looks like with Modern Tools? Like, so for instance, when a writer finishes a draft for a TV episode, who needs to sign up? Producer, showrunner, network, legal? And then how do tools facilitate that? Like, I'm curious where the bottlenecks usually are. Is it in tracking who is approved?

‍

What in communication delays or is it something else? And how do digital platforms streamline those approval chains that used to happen to your point over phone calls or emails?

‍

Josh Kline (13:30.476)

Yeah, so to be frank, some of that is TBD. Some of that we're building. Some of that we're getting feedback from our users. In the area of what internally is called studio TV, but the audience would think of as like variety shows or late night type formats. Generally speaking, if it's the late show with Stephen Colbert, the approver is Stephen Colbert, right?

‍

When the one when whether it's, you know, last week tonight with John Oliver or the Daily Show, the exec producer in some cases, the show runner person whose name is on the big door like they're going to approve it for network programming. It's usually an experienced show runner would be the one approving it. If you have like a John Wells or JJ Abrams, somebody that's stature.

‍

Greg Berlanti, like, you know, they understand what is going to be approved. If you have a sort of a younger, maybe less experienced showrunner on a project, they'll probably get more feedback from the studio or the network. So those systems aren't always online yet or even developed yet. So you may get a PDF of a script that gets sent to somebody for comments.

‍

And that's something I'm tackling with my team right now, which is this concept of if you have the data in a platform where you can securely share that information with other people, you think about it like a contract among lawyers. Like, I'm going to share access. You're going to look at this. You're going let me know if that's OK. And then we'll build workflows around it. So it's not always like digging into the document and making specific comments.

‍

It's review and approval workflows. And so we had those back in the day with digital dailies. You we were providing cuts and dailies for review and approval, but we also had workflows baked into that. an example in scripto would be, we've got a Slack integration and as you're making comments and you're sending it down the production pipeline, you know, I could, if you were gonna, if you were my production executive, I would make a comment and I would at camera in it.

‍

Josh Kline (15:53.644)

you'd get a Slack notification, and then you'd go right to that comment, you'd see the changes, and you'd say, I'm cool with this or I'm not. So then we'd have it in Slack, and we'll add additional platforms as we go. But today it's a Slack-based notification approval system, and you would make your comments, and we would know if it's approved or not. That's not the end game for a review and approval system. It is a way to document it, and it's referenceable.

‍

But it's not like an automated review and approval process. So we're building out a number of those right now. And it's all based on feedback from the users, right? We need to know what the showrunners, the exec producers, the executives, the financiers need out of a system like this. And to be clear, Scripto came from the variety space where they're putting out a show every day or once a week.

‍

at the sling last week tonight with John Oliver, obviously a weekly show, SNL Weekend Update every week, the Late Show every day. So writers who was in a script all day long pumping out a script every day, there's not a whole lot of time for review and approval and waiting for people, but Steven's one of those writers.

‍

Cameron (17:09.729)

‍

You know, what's interesting is to that point, you know, in a recent interview I saw with you, you mentioned that when writers make last minute changes, that there's a risk of things slipping through the cracks from the script to the teleprompter or rundown. Like to your point, things are changing constantly. Can you give an example of this problem and then how a platform like Scripto addresses it? I imagine on a live show that a tiny discrepancy can mean

‍

Josh Kline (17:30.349)

Yeah.

‍

Cameron (17:41.1)

that the host reads the wrong line or a queue is missed. So how do you design a system that catches these issues in real time? And have there been surprising challenges in getting every stakeholder, the writers, the tech crew, et cetera, to trust and rely on the system's notifications?

‍

Josh Kline (17:45.176)

Yeah.

‍

Josh Kline (18:00.096)

Yeah, so we're certainly more feature complete in the variety side because that's the background of the company. We built up the original slate. I'm the new CEO. So my predecessor built this company around a well-known set of workflows and requirements in the variety space. And so they catch stuff. This was designed to catch things in real time because it's not just the writers in this platform. There may be 10 writers.

‍

and 80 production staff members involved in the platform. So they're all in it every day. They're getting notified of things that happen. And in that side of the house, they are expected to pay attention to what's happening. So they're notified. The changes are all highlighted. They're putting out show every day. And we create what would be a narrative scripted side of production would be a breakdown.

‍

that the department heads would reference, usually ADs would work on those, but all the department heads are getting a breakdown for their specific requirements. On the variety side, that would be a rundown. So it's a run of show. And we're updating that as the script is being updated. So it's tined down to the second. And as long as you're in the system, it's your responsibility, if you're producing a show, to pay attention what the rundown is.

‍

And so a department head has a responsibility to be up to date on the information. The good news is it's not going to get lost in an email. It's not going to get stuck in an outbox, right? It's like, it's a cloud-based collaborative system. You're in it, and part of your job responsibility is to stay up to date. We'll notify you of changes, but you're expected to pay attention. In terms of how that sort of gets juxtaposed on narrative scripted workflows,

‍

That is an opportunity for us to work with partners across the ecosystem. So we're not going to build out all of these systems. are, I talk about Scripto as being an operating system for scripted media production that has a collaborative writing app at the head end of it that is very efficient at writing and getting data into this platform. But we're only one set of data. There's multiple data streams.

‍

Josh Kline (20:23.65)

that run through a production. So there's scheduling and budgeting data. There is a set of VFX data. There's physical production data. And a lot of that hangs off the script. So you can peg much of this to, there's a new character name. Well, that has to go to Clearances. somebody's got to make sure that character name doesn't exist in the city where the show's being shot. That could be confused. There's all these nuances.

‍

that like these little workflows that fire off of what the writer's doing. So by having the writing connected into the data stream that feeds all these downstream workflows, we get these real time updates, right? We get real time changes flow through the system, but it will be other partners that are responsible. So we're not going to be the company that makes sure that the clearances get done.

‍

Scripto is going to be the platform that creates notifications. And some of those are UI-based notifications, like Slack. But we're an API-enabled platform. We're firing off calls out of the API all the time. So that will include third parties and their systems. Some of this is going to flow into what you do for budgeting.

‍

And the finance side of the equation, changes will be made that impact the data you're managing. So some of it we handle. What's natural for us to handle on certain shows in certain formats makes sense for us to handle. And what's not, we're actively looking at doing integrations with third parties. And there's, you know, we talked recently about movie labs and the 2030 vision, and that's all about

‍

cloud-based, connected, interoperational workflows. And data streams that flow, regardless of which partner you actually plug into those flows. And an example of that would be like, I want to play around with an AI tool that gives me concept art. say there's like five of them you're kind of interested in seeing.

‍

Josh Kline (22:50.126)

what the output looks like. I'm going to make some updates in a script and scripto, and I'm going to do an API integration to Dolly first. But now I maybe want to check out a couple of different ones. I want to check out stability. I want to check out three or four other ones. It doesn't affect the data stream. You're just plugging in a different tool that's accessing the data. And so that's what this very efficient forward-looking workflow looks like for production is

‍

Cameron (23:17.016)

workflow looks like for production.

‍

Josh Kline (23:20.366)

partners that can access the data, they've got the security clearance to access the data. I own the IP, not me as Scripto, but as a studio or as a production company, I own this IP. I decide what tools I want to use. I'm going to use, Scripto as the platform that carries the valuable data and allows me to plug in whatever tool I want in a really efficient way.

‍

Cameron (23:45.9)

That's awesome. You know, you pioneered digital dailies to streamline how directors review footage. Do you see parallels between managing video dailies and managing scripts? I mean, I think you're talking about that with sort of the operability of data flows in our business, but I'm curious for you, like what lessons have carried over from building a dailies platform to building a script collaboration platform? And are there any new surprises unique to the script workflows in particular?

‍

Josh Kline (24:01.934)

Yeah.

‍

Cameron (24:14.06)

despite your earlier success with dailies.

‍

Josh Kline (24:18.326)

It's an interesting shift. Daily, we were, I mean, this was a while ago. So our competitor in Daily's was three quarter inch tape and DVDs. And my team just thought there's a better way to do this. Like, this is silly. Streaming media was coming online and we thought we can use these tools in a way.

‍

that modernizes this workflow. So a lot of the lessons we learned were about how to move people off of legacy platforms into more efficient modern workflows. But back 20 years ago, all sorts of stumbles along the way. Not everything worked perfectly. It was a high touch business. We're certainly dealing with lighter touch assets. Excuse me, these are, it's text.

‍

I'm dealing with text. I'm not moving terabyte files around. My AWS bill's a lot cheaper now than it was then. And we didn't even have AWS when we started. So I'm not building servers anymore, deploying them in data centers. So some of it's easier. The expectations around the quality of software. So you can't roll out something that's like OK. Nobody will use it.

‍

I'm not rushing version 0.1 kind of stuff to market. What I've learned along the way is I have the same expectations. We're in an era where software should be of a high quality. I'm trying to get people to move from one platform to another. It can't just be functionally superior. It has to be a more pleasing environment in which to work.

‍

Cameron (26:40.034)

How do you balance imposing a structured workflow with giving artists freedom? Or have you encountered pushback from writers or producers about using a tool for scripting? Maybe it's related to it has to be beautiful, or maybe it's feeling too restrictive or technical. And then how do you think about overcoming that? Essentially, how do you design collaboration software that's both rigorous on the back end without stifling creativity or usability on the front end?

‍

Josh Kline (27:08.46)

Yeah, so my balance is I know a lot of writers. In my world, I'm surrounded by a lot of writers and producers. And so I take a lot of feedback from creatives. And I certainly lived through the era of enterprise software was deployed by the CIO's organization and like everybody's going to use it because the CIO said so even if the consumer tool that you'd prefer to use is way more elegant.

‍

You have to use this stuff. So we're past that, right? Because people, this sort of thing we dealt with at Box, which was we go to an enterprise and we do an audit for the enterprise and say, you've got this shadow IT world living in your network. And here's like 100 different apps people are using because they'd rather use that stuff than what you're providing. So let's give them something they want to use. Otherwise, they're just going to avoid the stuff they don't. It's like trying to deploy, you know,

‍

enterprise social networking software like Yammer. I tried to implement Yammer at a company once and it was like they didn't want to do it. They wouldn't do it. They refused. And then Slack comes along and it's a really elegant solution and everybody wanted to do it. So the software speaks for itself. In what I do now, you know, I...

‍

We're not in a position, we don't have market dominance. We've got a solution that provides value for writers that need to collaborate. But writers that have no need to collaborate may very well say, yeah, I've been using Final Draft my whole career. I've got friends to say, I don't really need to learn anything new. I'm doing just fine and I don't collaborate with anybody. So when I'm done,

‍

I send the file to somebody, they have to figure this stuff out. So in a situation like that, next step is to work with the, they need to figure it out. So we work in certain instances, one step past the writer, one step further down the supply chain from the writer to these different people on a production team that are dealing with mayhem.

‍

Josh Kline (29:31.822)

Like the writer may think it's going fine because they're emailing a PDF file to a script coordinator and that coordinator is like, you know, I'll date myself with this one, but I don't know if you remember the video game Missile Command where it's like, you know, it's like a Cold War thought of like nuclear missiles coming in from every angle and you're trying to shoot them out of the sky. Like that's the analogy for a script coordinator. They're getting like

‍

PDF drafts of scripts all the time. They're trying to figure out is the naming convention telling me what draft this is so every show's different. They've got to figure out the diff between this version and that version, what they need to then send along either into another tool that's doing the distribution or many shows just do email distribution still. It's a lot of work. I'll say,

‍

On the record, I think that the script coordinators are the unsung heroes of the scripted media production industry. I really do. So we spend a lot of time working with them, then physical production teams, department heads, ADs, UPMs, line producers that depend on information that has changed from draft to draft. And so it's really like, it is discovering pain points along the way to figure out what we can do to help. We don't have

‍

market dominance in a way like nobody does like that can force people to do anything beautiful software is a solution for us and we you know for our variety show type clients I think most of them would say we deliver beautiful software for their needs and as we you know we've got customers in podcasting award shows streaming creators

‍

We don't yet have a narrative scripted commercial product. We've been building one since I, I mean, you can write a script in script out today and this formatting is all there, but it's really like having a solution that'll carry you into production. And there are just some specific tools within that format that we're building out and are soon to launch. But if we're incomplete, for instance, the reason we don't have a solution in market that we're monetizing is

‍

Josh Kline (31:57.292)

Because we're incomplete, we're feature incomplete. So if I rushed one to market, I think I will have shot myself in the foot. And when it comes to software and trying to deliver beautiful software for people, you get one chance to make a first impression. when I, you know, I think about our customers signing into Scripto for the first time, I want them to be delighted.

‍

Looking forward, how far do you think automation can go in production workflows? You know, can you envision in your future where say AI drafts scene and humans tweak it or where the system automatically flags budget issues in a script draft? You were mentioning that a little while ago. You know, what parts of the scripting and production processes do you believe will always need a human touch no matter how smart our tools get?

‍

So I think about maybe like a three-year event horizon around technology. What's going to happen in the next three years? that moves pretty fast. And anything past that is a little hard to see around the corner, at least in the area where I operate. And I'm seeing automation workflows that are pretty impressive.

‍

But what I think is that there's still a requirement for human intervention on almost all of them. Because I haven't seen any AI-powered workflows where people would say, I will bet my career on the output of this tool. So there are some, for instance, breakdown creation tools that are cool. They're great. But

‍

You know, a breakdown, say it's a six million an episode show and you're the UPM, the line producer, you're an AD and you're using one of these tools. Are you going to publish that breakdown to all the crew members to show up, you know, a couple of days from now to shoot this scene? Are you going to, is there a single tool or?

‍

existing piece of technology that you would trust enough to risk your career on. And I think resoundingly, the answer is no, I, I can't even imagine a working producer would say yes anywhere to that question. But what I do think is in a situation where somebody is using a platform like scripto, where a change is made to the script.

‍

that data flows through to other systems, one of which would be breakdown creation. And you're going to have an approved breakdown. You're going to have a script change. And now what you have is a diff. And you're going to get a highlighted piece of data. We've got a difference between these two breakdowns. And that's going to send a notification to an AD or somebody on the production team. And then

‍

Josh Kline (35:06.99)

Basically, you've cut out like 80 % of the work, right? You're not reading the script to see what's happening. You've been notified. It's been highlighted. Now you got to go in and you may get a suggestion as to how to update the breakdown. And you can, you know, that's like a automation initiated review and approval process, right? All that work happened. The writing part was manual. Once the draft was saved, everything else was automated. The updated breakdown.

‍

and now you have somebody whose job is on the line to look at those two files, or the UI between the data, and say, yeah, that updated breakdown is right. I approve. Or it's pretty good, but now I'm going to go in and fix it, and I've still saved hours. I think that's the benefit of the automated tools.

‍

Cameron (36:02.072)

You know what's interesting, Josh, I was going to ask you as my last question, you know, as a thought experiment, if you had no legacy constraints and could design the ideal production workflow from scratch, what would it look like? That sounds pretty good. I mean, are there any other radical changes you would make to how script schedules media is managed? You know, like if you could, you know, wave the magic wand, you know, what's the next big leap in workflow that you daydream about, even if it's not fully feasible today.

‍

Josh Kline (36:33.154)

Yeah, so I'm so focused on the data side and the text and the words. So I'm a step removed from the content generation side. So I'm honestly not the one to answer it on the content generation side. There are people that give such a better answer than I would at this point. But on the data side, it is really

‍

Like I believe wholeheartedly in the work that Movie Labs is doing for 2030 vision. I ascribe to that mindset about, you know, let's have uninterrupted data streams in the cloud. Let's have interoperable, interchangeable companies and tools that plug into that data stream. We essentially pick the one we want that makes our job easier. And those tools will get better and better over time.

‍

I like standards. I like standardization of data formats. So we get to this kind of Lego block way of, this will sound wrong, not producing content. We don't want cookie cutter content. But we get known and repeatable processes for how to create the environment that creates that content. And I think we're heading in that direction. So that's part of the excitement.

‍

from my side of what I'm doing right now is that I was invited to lead a company and a team that was already heading down that path. And I get to bring to bear 25 years of working in the industry where I get to sort of optimize around that.

‍

Cameron (38:17.228)

That's awesome. Well, Josh Kline, thank you for illuminating me and the listeners on what you're doing with crypto. It's really exciting and it's really excellent to chat with another technologist trying to drive our awesome entertainment industry forward. Thanks for being with me.

‍

Josh Kline (38:32.632)

Well, I thank you very much for having me. It's always a pleasure to talk with you and I appreciate the time.