Today, we sit down and talk with Ross Belmont, the Director of Product Management at Salesforce, about his products – Salesforce Connect and the new DynamoDB Connector.
Salesforce Connect provides data virtualization, surfacing structured data inside the Salesforce UI and declarative tools experiences, even though data resides outside the four walls of Salesforce. Their vision is to get to a point where it no longer matters that external objects are external.
Ross also shares the current status of DynamoDB coming in pilot. Their team is currently working with Amazon Web Services, extending Salesforce Connect ideas over to Amazon’s DynamoDB, a super high-volume, NoSQL datastore.
Show Highlights:
- Ross’s first-ever computer
- His entry into programming and product design
- What enables Salesforce Connect?
- The role of OData
- Other examples of entities Salesforce Connect is attached to
- Our current relationship with Amazon: why and how are we working with them?
- Challenges around DynamoDB
- Authentication and decomposing credentials for more flexibility
- The evolution of world identity and access management
- Examples of standard use cases for SQL
Links:
Connect with Ross Belmont on these platforms:
Episode Transcript
Ross Belmont:
Sort of try to do what you can to pave the road in front of them so that your developers can come through as efficiently as possible and make progress because their time is super crazy valuable.
Josh Birk:
That is Ross Belmont, director of product management here at Salesforce. I’m Josh Birk, your host for the Salesforce developer podcast and here on the podcast, you’ll hear stories and insights from developers for developers. Today, we sit down and talk with Ross about his product Salesforce Connect and the new dynamo DP connector that’s coming into a pilot. But first we start as we often do with our early years with Ross’s first ever computer.
Ross Belmont:
Oh boy, that is a good one. So I’ve had trouble nailing down what the precise name of this product was, but I think there was an Apple II clone that was built and marketed by a company Laser. I swear to God, it was embossed onto the plastic of the casing.
Josh Birk:
Really?
Ross Belmont:
Yeah. And now I could be imagining things, but we did have a unit like this at home. So I had some kind of Apple II at home before getting our first Mac. I want to say it was an LC2.
Josh Birk:
Okay.
Ross Belmont:
And maybe 91 ish or something like that was the first sort of real quote unquote computer we had as a family. And I immediately sort of broke it by putting on all sorts of system extensions that made it effectively unusable in terms of its ability to start the operating system. So those are good lessons to learn.
Josh Birk:
Wait, roll back there. What kind of system extensions are you putting onto an old school Mac that’s going to say, it just gave up trying to boot?
Ross Belmont:
So it was the days when each extension you would load would add a puzzle piece to the corner of the loading screen.
Josh Birk:
Right. Yes.
Ross Belmont:
And it would start to have multiple rows of these extensions, like expected people to have two or three or four or five.
Josh Birk:
Yeah.
Ross Belmont:
But I had like three rows of these things. And it was just taking forever to load and my dad, who had bought the computer because he worked in a school [crosstalk 00:02:19] at the educational pricing was like, “What is happening here? What did you do exactly?” Those were good times. And then shortly thereafter, he got me my own computer, as you can imagine, [crosstalk 00:02:34] just to sort of contain the damage, which made sense.
Josh Birk:
Nice. Nice. Was that also your first entry into programming?
Ross Belmont:
Let’s see. No, I mean, programming on those early Macs was pretty nasty. Like you had to work. I think there was a way to work in Pascal, but most of it was-
Josh Birk:
That’s right.
Ross Belmont:
In C. I did in, let’s see high school, I did an independent study with a teacher. There was no computer class or programming class in the school but I did an independent study with the math teacher and we used a language called FutureBasic, which was a third party thing, that you could write basic style stuff for the Mac. It would give you a basic UI. It was not exactly like a Visual Basic, but sort of kind of close.
Josh Birk:
Gotcha.
Ross Belmont:
And I made a calculator that was very embarrassing. That’s kind of what I remember taking away from it, but it was good to mess around and write some truly horrific code.
Josh Birk:
I mean, embarrassing in terms of like UI because it’s a calculator, I’m assuming it at least knew how to add and subtract and stuff.
Ross Belmont:
It did. It could do the math, which was great. Since the run time had all the math functions built in, then what I did is just, I went nuts and I added like 30 items to every menu so that you could do the arc tangent and all those things. Like imagine like every button on a scientific calculator, but just represented like on screen. So there was not a lot of elegance and restraint in the UI at the time, but there was a lot of gusto.
Josh Birk:
I’m just going to throw this question out there, just kind of out of the blue. Throughout your career, have you ever had any trouble with scope creep?
Ross Belmont:
Well, that is a great… I would like to say that I learned my lesson, didn’t I? It was like, hmm, let’s take a look and see, do I really need this arc sin capability in one O, or can we squeeze down the MVP a little bit and see if we can get to market a little sooner, but…
Josh Birk:
Gotcha.
Ross Belmont:
Took a while to learn that lesson.
Josh Birk:
Okay. That’s actually interesting because I did have a question on the list was, which was kind of like, when did you start moving from like just kind of purely being a coder and programming and tinkering [inaudible 00:04:58] stuff, to thinking more about like product designs and things that people might actually buy and it sounds like that was your earlier years too?
Ross Belmont:
Yeah. Well, I had thought about it then and it took me a while to bounce around and because I was looking at the different roles that you had on these teams and that’s changed a lot, as you can imagine since 1994, but the thing I was always trying to figure out, even though I didn’t really know was what does it mean to get in leadership position for this, to be able to decide what the feature set’s going to be.
Ross Belmont:
And it took me some time actually to eventually figure out that’s what we now call product management. But at the time you would have just sort of like the tech leader that was maybe more able to communicate, would naturally take that role. Or then I thought for a while that it’s all about the design leadership, but was looking at companies like Apple. Or I remember, especially like early Dropbox, really design driven companies like that, thinking about that approach and then eventually getting enough sort of global knowledge of the industry to figure out that there was a job for that, at least in the situation where the company makes the products for sale. If the company’s not making the products for sale and you’re just in IT or in consulting, then it obviously gets a little bit muddy as to who’s really going to make that call.
Josh Birk:
Right.
Ross Belmont:
But I spent a while in that area as well sort of bopping around the different leadership roles saying, “Should I do this? Should I do this? Should I do this?” And then eventually, some years later made my way over to product.
Josh Birk:
Got it. Got it. And then what was your first introduction to Salesforce?
Ross Belmont:
Oh yeah. So I was in IT at a mid-sized accounting company and the person who was in charge of the data group, like the enterprise data warehousing and analytics team, we were on a project together trying to put together an early version of something like a warehouse and thinking about having APIs around some of the sort of core data that you might now think of as like a data lake but before people were really using that term.
Josh Birk:
Yeah.
Ross Belmont:
He was saying like, “Wow, I wish we had Salesforce’s APIs, everything they do has really good APIs sort of automatically.”
Josh Birk:
Yeah.
Ross Belmont:
I was like, “And tell me more about this Salesforce thing. That’s really interesting.” Before then, I had done some Java and then I could sort of recognize the Java roots in Salesforce. And so I immediately sort of got comfortable with it and then I thought, “Okay, maybe I’ll do this instead. I don’t really know how I feel about this SharePoint business everybody’s telling me about here in IT. So let me go try that instead.” And not long after that was probably around the time that we first met actually, way back 2010ish timeframe.
Josh Birk:
Yeah.
Ross Belmont:
And then there you go. So I spent a lot of time after that doing apps on the app exchange.
Josh Birk:
Nice, nice. That was my evil developer advocate laugh by the way. [crosstalk 00:08:12] Whenever I hear SharePoint, it’s kind of just sort of rumbles out a little bit.
Ross Belmont:
Yeah. That’s warranted, I think. Yeah. I think that’s totally reasonable.
Josh Birk:
Nice. And then how did you end up at Salesforce?
Ross Belmont:
Oh, okay. Well, so, in that time that I was doing the app exchange apps, I had met a whole bunch of folks on the ISV team, both sort of tech evangelists and the business side folks and some of the product side folks and one of the folks who was in one of those roles had moved over into product and platform, where I am now. And it was sort of just luck that she had a situation where her area grew. She had a number of dev teams to manage, maybe a few too many and then they were able to carve out a space for somebody like me to grab a few of them.
Josh Birk:
Nice.
Ross Belmont:
And sort of water that little patch of grass, as they say, right over in platform.
Josh Birk:
Nice.
Ross Belmont:
Here I am.
Josh Birk:
Okay. So I know it’s sort of right on the tin, but describe Salesforce Connect to me, like what is the current product and what does it do?
Ross Belmont:
Sure. So, Salesforce Connect provides what we like to call data virtualization, which is surfacing structured data inside the Salesforce sort of UI declarative tools experiences, even though that data resides outside the four walls of Salesforce and outside our regular data store.
Josh Birk:
Mm-hmm (affirmative).
Ross Belmont:
This is becoming more and more commonplace and popular.
Josh Birk:
Mm-hmm (affirmative).
Ross Belmont:
And probably the world is getting more polyglot than less over time.
Josh Birk:
Yeah.
Ross Belmont:
So the need for this type of thing continues a pace and we enable that functionality by what we call these different adapters. Each adapter connects out to a given data source that is of a certain type and maybe there’s a certain protocol or data format that data source then supports and Salesforce Connect will translate that to a general structure that lights up everything that admins know and love about the Salesforce tools experience. Everything from list views, to flows, to reports, apex, and everywhere in between.
Josh Birk:
Mm-hmm (affirmative). Would you say it’s safe to say the goal is for developers and admins to be able to work and treat with external objects, just like they would standard or a customer?
Ross Belmont:
Yes. That is a sort of vision goal for our team, is to get to a point where it doesn’t really matter that external objects are external.
Josh Birk:
Got it. When I first started evangelizing Salesforce Connect, I know there was a lot of talk around the OData protocol and how we were kind of leading to OData one and then kind of waiting for OData two and then we leaned into OData two. First of all, for the listeners not familiar, could you define OData for us?
Ross Belmont:
Yeah, sure. It’s sort of a protocol/standard. It is a standard. There’s an industry group that looks over it like any other standard, like HTTP or anything else, largely spearheaded by the folks at Microsoft and SAP.
Josh Birk:
Got it.
Ross Belmont:
Who want better ways to get data in and out of their respective data sources, possibly for different reasons but this is really advantageous for legacy data sources that would be difficult to access in their sort of raw tab tabular format. If you were to try to directly load to take a look at the SAP tables, you’d be disappointed probably.
Josh Birk:
[inaudible 00:12:07] You might be a little confused.
Ross Belmont:
Yeah. In their direct business applicability for your average end-user, right?
Josh Birk:
Right.
Ross Belmont:
So they’ve done a lot to put the layer in front of it that provides a nice rest style API to get to that data, but not only that, it’s more than just what you would think of as maybe a typical rest structure that allows you to access a record and use a different HTTP verbs. They also get into advanced querying that would cross different types of objects, that would be like a joint equivalent.
Josh Birk:
Okay.
Ross Belmont:
And filtering and things like that. So the OData takes those ideas to the next level and it’s very mature in terms of what it offers around all that.
Josh Birk:
Got it. What are some other examples of entities that Salesforce Connect can currently attach to?
Ross Belmont:
So there’s OData, which we just mentioned. We have some other adapters as well. So we have a cross board adapter, to go to other Salesforce orgs, different sort of story around that, because obviously we have a lot more control because we know what both sides of the equation are.
Josh Birk:
Right.
Ross Belmont:
That has kind of a different evolutionary path which is really interesting. Then we’ve got an apex adapter, which can let you write apex code to access any other data source with any API and surface its data in the same way again, as if it were an object. So this is pretty nifty.
Josh Birk:
Gotcha.
Ross Belmont:
Yeah. So you can write a little apex. It’s not too tricky, but it is doable. There’s an interface, basically you implement against the apex layer and then that external data shows up as objects and lists and things like that.
Josh Birk:
Neat. So like, I think I missed that in the doc. So I could write a little apex that would surface say Twitter data that I’m getting as JSON, but the system could see it as an external [inaudible 00:14:09].
Ross Belmont:
Right. You might have, then you might use that to define the tweets tab, so to speak and then things like that. And then it also includes support for metadata, which is really fun. So if you can look into that other system, understand its metadata, send that back through, then the objects and fields can be created automatically.
Josh Birk:
Got it. Neat. Okay. So let’s talk future stuff, I guess it’s for near future stuff and a little bit of current stuff. So actually describe our current relationship with Amazon and why and how are we working with them.
Ross Belmont:
Yes, exactly. So it was last summer that a big announcement went out about the Amazon partnership. Specifically relative to AWS and making investments at first, the platform layer.
Josh Birk:
Yeah.
Ross Belmont:
So that our platform and theirs can work better together in a variety of ways that you might want, if you are a customer on both sides, and then the evolution of that will be later on new applications built, definitely on the Salesforce side, with leverage interesting tech that AWS has, that we may not on the Salesforce side today. Right?
Josh Birk:
Right.
Ross Belmont:
We have to kind of go first from the platform perspective so that any app sitting on top can then take advantage. Right?
Josh Birk:
Got it.
Ross Belmont:
So what we’re doing on my side of the house with that right now is extending this Salesforce Connect ideas, everything we’ve talked about thus far over to Amazon’s DynamoDB, which is a really interesting, super high volume, no SQL data store they have that’s very popular in their world. And there’s other things we’re doing eventually, but that’s the one that’s in pilot now that we’re really excited about.
Josh Birk:
Got it. So when you say high volume data, what kind of use cases are you seeing there, what kind of data are we putting in?
Ross Belmont:
Sure. Well, I mean, in general, just broadly speaking, legend has it, that DynamoDB itself was invented to handle the e-com store traffic that we all know and love that makes up Amazon from the consumer point of view.
Josh Birk:
Oh.
Ross Belmont:
Right?
Josh Birk:
Okay.
Ross Belmont:
So your orders and the status of your orders and the shipments and everything in those orders and all that under the hood is stored in Dynamo and they were reaching the limits of what was doable with relational databases. So the story goes to get the Black Friday level throughputs that they need to be able to support.
Josh Birk:
Right. Gotcha. Gotcha.
Ross Belmont:
Yeah. Now that’s that, and I know most folks don’t have maybe quite those same workloads, but what we do see are, we see things where there are lots of records that are maybe transaction logs. Those are interesting use cases I’m starting to hear about.
Josh Birk:
Okay.
Ross Belmont:
Every change to every part of a customer order in a B2B situation, that’s heavily audited. And maybe you don’t need or directly want to have that inside the fort walls of Salesforce, but you want to be able to look it up. Right?
Josh Birk:
Right.
Ross Belmont:
Or some something that may not be in CRM, like all the shipping event status changes on your orders. Right?
Josh Birk:
Right.
Ross Belmont:
Things like that you would want to know.
Josh Birk:
Yeah.
Ross Belmont:
There are things that we’re seeing, there’s a lot of commerce related use cases, extending sort of deepening around everything that you can think of that might attach to an order. I think that’ll be relatively popular and we’ll see how that goes.
Josh Birk:
Got it, because there’s a lot of times when we’re looking at the big flow chart of responsibility, like the Salesforce CRM is just the eye on the [inaudible 00:18:09] chart. We just need access to that information, but it’s the other guy who’s actually putting the stuff on trucks, right?
Ross Belmont:
Yes. Yeah. I mean, I don’t know. I don’t think there is a shipment standard object or a shipping app that will sell you quite yet. Right?
Josh Birk:
Right.
Ross Belmont:
We do have order management as an extension to Service Cloud, but eventually. The further you get into maybe traditional ERP territory, inventory, what’s going on in the warehouse, etc and so forth. Right?
Josh Birk:
Yeah.
Ross Belmont:
The more likely it is to be somewhere else, but your sellers and service agents probably want to see that data at some point when the customers call into the call center.
Josh Birk:
Right. And they want quick because this style of database is deep and it’s fast, but it’s not terribly fancy. What are some challenges with talking to this kind of data structure?
Ross Belmont:
Yeah. So to achieve the crazy high B2C scale that Dynamo in particular is capable of, there are no such things as joins. So you change your approach to data modeling and data storage to match the data access patterns and if you take a look at the Dynamo guides on how to think about these things, they really very strongly encourage everybody to understand upfront the different sort of access patterns that your apps need to support.
Josh Birk:
Okay.
Ross Belmont:
Almost in the same way as imagine if you knew the reports first for a given Salesforce app, that would effectively translate down into queries basically at the end of the day, but you would know how you’d be sorting them, grouping them, et cetera. Right?
Josh Birk:
Okay.
Ross Belmont:
If you don’t know that upfront, you probably won’t get the results that you would like later on down the line, because you’ll be in a situation where you look at a need and your query becomes a full table scan and that’s not where you want to be. So you have to know all that upfront.
Josh Birk:
Got it. Because that was kind of going to be my follow up question is like, where so on there’s the data structure part of it, but are there things that like the apex developer should consider when acting against, it’s going to act like an external object to them, but it’s really talking to DynamoDB, just so is there dos and do nots there?
Ross Belmont:
Yeah. I think it’s more of the cascading effect of the fact that you won’t be able to write absolutely any query under the sun that you can think of in the moment that you didn’t plan for upfront. Right?
Josh Birk:
Got it.
Ross Belmont:
Relational databases basically give you exactly that freedom and if you need that, you should probably look elsewhere. Right?
Josh Birk:
Okay.
Ross Belmont:
But you know though that the trade off is eventually difficult performance characteristics at the really high end of volume. Low and medium traffic is fine, really high traffic becomes really painful and so you look at the DynamoDB resources that are online and their AWS reinvent talks and things like that. I have this clear picture in my mind of these graphs where this performance is consistent, it’s a straight line. As usage goes up, performance is always excellent no matter what, doesn’t matter. As volume goes up, it’s in constant time, so to speak, right?
Josh Birk:
Got it.
Ross Belmont:
That’s the magic, but to that, you have to model to that and know how you’re going to need to grab it later. Right.
Josh Birk:
Got it.
Ross Belmont:
And then it’ll work.
Josh Birk:
Got it. So it does a little magic, but you have to play by the rules to get the little magic.
Ross Belmont:
Yeah, exactly. Right. But then when you do that, you can pull up your order history among hundreds of millions of people who have made trillions of orders and that can work in subsecond result times.
Josh Birk:
Got it.
Ross Belmont:
That part is nice.
Josh Birk:
That’s useful. That could be useful. Yeah. We’ve been talking a lot about read. Is there distinctions with the right side of things?
Ross Belmont:
Actually, that there’s a good story there in terms of Dynamo. So dynamo has great APIs for us on both sides. So edits made, I mean, the Salesforce, Y or [inaudible 00:22:39] apex or in a flow or something like that can usually sail in just fine. That’s not typically a problem. Like if you have a really strongly opinionated sort of application layer that treats the database maybe in an interesting way, then possibly that direct right could confuse your app.
Josh Birk:
Got it.
Ross Belmont:
If you’re not in that situation, which I actually don’t know how common that would really be, it’s theoretically possible, but what we’re understanding is that’s probably not super common.
Josh Birk:
Got it.
Ross Belmont:
So I think the read-write capability will be in good shape. I think the main limits and things that people will end up keeping in mind from the Salesforce perspective is again that full reporting that you’re all used to, not every aspect of that will work as smoothly as you’d like if the data isn’t modeled in a way those queries will work.
Josh Birk:
Got it. And I guess this is even just kind of a broad question. How is all of this handling authentication?
Ross Belmont:
Oh, sure. So I happen to also own the name credentials aspect of platform, and we’re doing a lot of work in that area to not only bolster the support that we have already for AWS’s signature V4 protocol. Right?
Josh Birk:
Okay.
Ross Belmont:
But to expand what we offer directly and how name credentials is, exposes that for Salesforce admins and devs to provide them more.
Josh Birk:
Got it.
Ross Belmont:
That gets into a little bit of a different story, but the future of that is really such that in today’s world, name credentials is monolithic in the sense that sort of endpoints and the protocol and the definitions of the actual credentials themselves are all sort of glued together as one unit, as one entity.
Josh Birk:
Yeah.
Ross Belmont:
We’re going to decompose that to give you more flexibility, especially as an app developer and it’ll feel more like a credentialing framework that you can do different things with, if need be, if you have advanced use cases. So one, I think easy way to understand it is, there will be lots of different applications and services in most cases running on AWS for a given customer. But I don’t know that it’s the truth that even though you might have say a dozen services that you’re relying on with AWS.
Josh Birk:
Yeah.
Ross Belmont:
You probably don’t want to make a dozen different credentials inside Salesforce. So we to split up what we have now and sort of the identification of the named system on the other side will be split from the credential itself. So the credential can be shared between multiple apps if need be. So connect to Dynamo over here and also EventBridge.
Josh Birk:
Got it.
Ross Belmont:
Or your Lambdas or whatever else.
Josh Birk:
Got it. That goes kind of into talks I’ve had with other people about like there’s authentication and then there’s also almost like your digital identity. It’s like this is me in this space and allow me access to these places that me in this space would need in order to get my work done.
Ross Belmont:
Yes. And it also leads to what the future of that technology is on the Amazon side, both in how they have the signature V4, sort of like that the cryptographic signing of a single request is one thing, but also the way that they have their IAM structure, their equivalent of profiles and permission sets. Right?
Josh Birk:
Okay.
Ross Belmont:
That’s what they call it in their world, identity and access management.
Josh Birk:
Got it.
Ross Belmont:
Those two work hand in hand as you would imagine, and those are evolving. And then as those infrastructure pieces are able to do more from sort of the server side of it, us on the client side, so to speak at Salesforce, well, we’re going to be developing some pieces hand in hand so that we can strengthen the security and make that easier to manage and things like that as both of those evolve together. Right?
Josh Birk:
Got it.
Ross Belmont:
Yeah. Some good stuff happening there too.
Josh Birk:
Nice. What’s the current status of the DynamoDB connector? Is it an open pilot, what kind of status is it?
Ross Belmont:
Yes, exactly. So it is in pilot. So you contact your account rep to then go through the process of the nomination on our side.
Josh Birk:
Okay.
Ross Belmont:
But I have a very expansive openhearted view come on in and join the pilot and let’s talk about your use cases and then let’s get you signed up.
Josh Birk:
Got it.
Ross Belmont:
Just more practical point of view, the way you do that is getting in touch with your account rep and then we’ll make its way over to me.
Josh Birk:
Okay.
Ross Belmont:
And we’ll get you set up in a sandbox so you can kick the tires.
Josh Birk:
And how long do you think you’ll run the pilot?
Ross Belmont:
We’ll probably do that for the scale of this release or the one that we’re in now. So then when we hit summer, we go GA with Dynamo. With the basic capabilities that we have now, what we’re doing is adding extra sort of like robustness and things like various limits we might need to put on, things like that tend to happen at Salesforce before you go GA even though most of the core functionality you’d see and be interested in it from the customer point of view was up on its feet for the pilot. Right?
Josh Birk:
Right.
Ross Belmont:
It’s that other productionizing that we’re doing now.
Josh Birk:
Yeah.
Ross Belmont:
So yeah. So the next few months basically is the time to get in and let us know. And then even that having been said, as these things are still released, we’re still ears open, taking feedback, listening, and evolving, both our side of it and us sharing that information back with the Dynamo product team and Amazon to make changes on their side and again, evolve together.
Josh Birk:
Yeah. Anything on the roadmap or what you would love to build in the future that you’re comfortable talking about, Safe Harbor [inaudible 00:28:46] looking statement [crosstalk 00:28:46]
Ross Belmont:
Safe Harbor, all that stuff, it’s an extremely Safe Harbor. There’s no waves, no, not a cloud in the sky.
Josh Birk:
Right.
Ross Belmont:
There’s plenty of buoys, no sharks. What’ll come after Dynamo actually at the same release timeframe, is another pilot of Amazon’s Athena query service, which is capable of querying almost anything on the Amazon side, but most popularly sort of known for its ability to query structured data sitting in S3. So really big CSV files or many tiny CSV files. Athena does the magic of sort of tableizing them, [crosstalk 00:29:32] so to speak, to coin a phrase, and then you can issue SQL directly to that and it brings back structured results. So that’s a nice, smooth bridge for us to walk across to basically translate cycle to SQL and we should have really awesome reporting type support for that adaptor when we get to that. So that’ll be really exciting.
Josh Birk:
What’s kind of a standard use case for that? Is it just sort of like, I have a bunch of legacy data, it’s all in the CSV file because that’s what got generated and boy, wouldn’t it just be nice if I didn’t have to convert it into a database and I could just treat it like that?
Ross Belmont:
That’s one way to go. Some of the people I’ve talked to already are taking the S3 approaches, basically the store for their data lakes.
Josh Birk:
Okay.
Ross Belmont:
And then you can do the SQL querying right on top of that. Amazon actually offers a visualization dashboarding tool that sits in front of Athena that does the queries. But I think the Tableau CRM and tools like that also know how to do that. The Tableau, I think Tableau and Tableau CRM both do that sort of thing. So if you like the storage costs of S3 and then you can have an awful lot of data over there pretty cheaply, especially if you do clever things with the raw files, like you can do sort of these different… You can avoid CSV and use these more optimized file formats that increase speed and things like that and decrease cost.
Josh Birk:
Got it.
Ross Belmont:
I think those might be the first use cases actually we come online with, in terms of what customers actually do and want to do.
Josh Birk:
And that’s our show, now at the time of the release of this episode, that pilot is still going on. So if you’re interested, please contact your account executive and give Ross some feedback. Before we go, I did ask after, Ross’s favorite nontechnical hobby.
Ross Belmont:
Oh, my favorite nontechnical… Boy, that’s a tough one. Probably I’m going to go with playing the guitar.
Josh Birk:
Nice.
Ross Belmont:
Done that for quite some time.
Josh Birk:
When did you start?
Ross Belmont:
Oh boy, that’s a good one. So I started that in college when it was really not workable to bring the drums. I played the drums earlier. It’s not really so much doable in the dorms.
Josh Birk:
No. Yeah.
Ross Belmont:
That’s kind of a no go.
Josh Birk:
Yeah.
Ross Belmont:
So, I ended up converting over to guitar and sort of stuck with that largely ever since. So it was just man, it’s just only like probably 72 or three years ago that I started with that, since the beginning of college, but yeah. But I keep them out. I keep them next to the chairs and the family room and things like that. So I can reach over and grab one when need be.
Josh Birk:
And just chill and or jam as needed.
Ross Belmont:
Mm-hmm (affirmative). As appropriate.
Josh Birk:
I want to thank Ross for the great conversation and information and as always, I want to thank you for listening. Now, if you want to learn more about this show, head on over to developer.salesforce.com/podcast, where you can hear old episodes, see the show notes and have links to your favorite podcast service. Thanks again, everybody and I’ll talk to you next week.