Launch HN: Taloflow (YC W21) – Find the top cloud or dev tools for your use case

[cloudfalcon]

Hi HN,

We are Jason, Todd and LV - we're building Taloflow (https://www.taloflow.ai) to help developers find and evaluate cloud and API products based on their use case, requirements, and budget.

Have you ever wondered whether you should move a particular workload off of AWS to GCP? Or which database to build on? Or which APM is best for your size, architecture, and budget?

A little over a year ago, we launched a cloud management tool for AWS. A few months after launch, we learned that many of the dev teams we were working with were actively looking for alternative cloud products and dev tools to improve their stack. They generally found the investigation and testing process to be error-prone and a time suck (finding what fits the use case requirements, reading docs, sitting on sales calls, projecting costs in complex spreadsheets, etc.).

Similarly, while building our own product, we implemented several cloud and dev tools into our own stack. Some worked, and some were terrible. Time after time, we were fooled by the marketing language, lacked use case-based information to inform our decision, and generally found third-party review sites not as helpful when it comes to buying a cloud or dev tool. Why? Comparing options with info relevant to our specific use case was the missing piece, and there were several dozen dimensions to consider.

We built two ways to help you pinpoint which products best fit your use case.

The first walks you through a "Quick match" questionnaire that is designed with the help of domain experts for particular categories (e.g.: object storage, data pipelines, APM, etc.) to help you define your requirements and then rank options within a product category based on the inputs.

For example, in object storage, we collect use case info (Do you provide information to customers frequently and quickly? Do you perform intensive ML tasks? Do you work with large graphical or video objects?, etc.), integration info (CDN, Redshift/BigQuery/ or other Data Abstraction Tools), budget, compliance, etc. to build a holistic use case profile to help filter and rank products.

We then scour our database of developer docs, pricing pages, private pricing, reviews from experts, and internal tests for various types of use cases to find the best use case fit.

The second way we solve this is by actually ingesting telemetry from various cloud applications (AWS, GCP, Azure, Twilio, Datadog, etc.) and providing a cost-benefit analysis on the performance and cost trade-offs of implementing various products. We can answer questions like "What are the egress and hidden fees of moving from Azure Blob to Backblaze B2, Storj or Wasabi?" or "What's the most performant APM for my architecture?", etc.

We’re currently experimenting with pricing, but we make money by selling more advanced or ongoing analysis to our members, up-selling our SaaS solutions for cloud infra and dev tool management, and occasionally collecting a standardized referral fee from vendors who happen to be good matches for you.

We're currently live for Object Storage and will soon launch for AI/ML tools, CPaaS (e.g.: Twilio, MessageBird, etc.), CI/CD tools (e.g.: Jenkins, Buildkite), and APM (e.g.: DataDog). If you don't mind getting ad-hoc analysis for categories we don't fully support yet, we have an option for that too. We'd love your feedback on our product and want to hear from you what product categories you need help researching.

[bradhilton]

I just tried out the tool for Object Storage and was pleasantly surprised with the resulting experience. I really like this; great work!

[omarhaneef]

It would be useful to see this tool used for other SaaS.

Want to manage a (1) 10-50 person (2) construction firm in (3) the US, should you use Basecamp, Asana or Wrike?

You have a (1) 50-100 person (2) Tech firm should you use docusign or hello for document management?

And if you already have and are committed to Quickbooks, does it change any of the answers above?

[akudha]

How would one go about building such a product though? Comparing hardware features (say iPhone with Samsung for example) is a lot easier than software, isn't it? SaaS products can release features every week, it would be quite hard to keep up.

It would be cool if there is a JSON standard to publish features (and pricing) for SaaS apps - something like what AWS does for its products. Defining such a standard itself would be tricky, but it is worth trying. It might make it a bit easier to compare various project management apps, for example.

[jasonkimtech]

Hey omarhaneef that's great feedback.

With regards to SaaS besides project management and document management what are some categories where you find picking tools the most cumbersome?

[omarhaneef]

I have never met any SaaS where it wasn't cumbersome to figure out the competitors, the differences and the pricing.

I think the greatest area that a small company could use help in is learning which tools are out there i.e. answering this very question. For instance, one firm I know had no idea how useful bill.com was and it saved them 10s of manhours per week but began using it years after it was offered.

The other area is workflow: you use software X and software Y in this way, but did you know most other companies link them up with software Z in the middle and it works much better, so long as you check the box that reminds you to tag the invoices (or whatever).

[jasonkimtech]

>The other area is workflow: you use software X and software Y in this way, but did you know most other companies link them up with software Z in the middle and it works much better, so long as you check the box that reminds you to tag the invoices (or whatever).

I think adding up all the time savings for certain workflows or integrations across different sets of products would be immensely useful. You're totally right there's no easy way to see that currently anywhere.

[omarhaneef]

It would be valuable just to tell people what others are using. The most popular configurations would lead to suggestions based on usage in the field.

This is sort of the reason people check out state of Javascript and learn that "everyone else" is gravitating towards Vue, or that Svelte is gaining in popularity, or that VS code is everyone's preferred editor.

[jcims]

Not the GP and have been out of the game a bit but CRM products were a nightmare the few times I had to select one. Security products are tricky as well.

[fthead9]

I think you're solving a problem that is becoming increasingly critical for many businesses. AWS, Google Cloud, and Azure are so enterprise focused that startups to mid-level businesses are getting killed once their storage credits run out and they have to pay full enterprise prices. Finding alternatives is a painful process of trial and error. Having a trusted recommendation engine would go a long way towards relieving that pain point. Good luck!

[flotothemoon]

Neat! I often find myself cross-referencing dozens of blogs, HN and Reddit articles to figure out all the little critical differences between products. Particularly in deployment & machine learning related software it's often hard to discern what the solutions actually provide beneath all the buzzwords on the landing page.

PS: The "Guide -> Select a category" page gives me some errors and requires a few reloads to work properly. I get: "Firefox can’t establish a connection to the server at wss://app.papercups.io/socket/websocket?vsn=2.0.0." (probably unrelated) "Cross-Origin Request Blocked: The Same Origin Policy disallows reading the remote resource at https://o454487.ingest.sentry.io/api/5636106/envelope/?sentr.... (Reason: CORS request did not succeed)" (also probably unrelated, but means you probably won't have this particular problem logged)

[stets]

Awesome to see this on the front-page, guys. Nice!

- if you monetize partly by collecting referral fees, do you plan to stay vendor-agnostic? how?

- over time, as products change via acquisition, updates, etc, how can you keep the various metrics, fees, features organized and up-to-date? This seems like immense work to me

[jasonkimtech]

Thanks!

We're currently experimenting with pricing (charging retainer or SaaS fees for advanced/ongoing analysis) but think we can keep referrals vendor-agnostic by standardizing referral rates and aligning incentives in a way where we get a small percentage over a long time period of usage. We're trying to optimize for long-term satisfaction and usage of a product so only want to be rewarded when we can fulfill that.

Also we'll always give a transparent reason for why each vendor was recommended or excluded based on your use case requirements (this vendor doesn't fit your compliance or cost requirements, this vendor has low average ease-of-use rating and you said ease of use was a high priority for you, etc).

In terms of keeping track of every product, feature, release - it's currently being done in a combination of monitoring tools for public docs, websites, and releases and in some cases we will directly communicate with the vendor to validate. In the future we're hoping to create a system of record whether that's an API or interface so vendors can push information to us directly (which we will then vet for accuracy) to keep things accurate and up to date. Eventually we want to help standardize terminology, and potentially even standardize some benchmarks so developers can view features and performance in a more uniform way.

[dkcdkc]

So dope to see y'all on the front page. Great team and product. Can't wait to see the future growth!

[frenchie4111]

There appears to be a recent trend of YC companies being increasingly "meta" plays. The increasing number of "startups for startups" worries me, it's a leverage against the success of startups. It will increase the upward and downward slope of the startup market. Leverage leads to bubbles, and bubbles burst

Note: This product looks great, I will probably use it, my comment is not intended to disparage taloflow specifically.

[ahstilde]

I'd actually expect this product to have more traction at larger companies who require executives to make purchasing decisions. Instead of farming it out to an outside consultancy for hundreds of thousands, here's a simple SaaS that will enable your engineers to make the best decision at any given time.

[zapita]

“startups for startups” is not so much leverage as it is a way to enter the market. If your competition is too entrenched to take on directly, you find a way to “wrap” it with an adjacent product. Then once you have a foothold with joint customers, you expand your product to until your complement is reduced to unprofitable commodity, and customers derive all the value from you.

[captn3m0]

SaaS companies for other startups has always been YC's core strength - I don't think this is anything new.

[_xnmw]

It's actually quite ridiculous, I actually thought YC is starting to feel like a pyramid scheme. I took a look at the recent batches of YC companies and more than 70% of them seemed to be of the "meta" type, startups selling to startups. Of course these companies do well initially being able to tap into YC's huge community of startups. That doesn't mean they provide fundamental value to the economy, they're just sloshing around a lot of VC money.

On the other hand, maybe that's intentional, kindof of how sports players practice on each other before facing others in a real game. YC is like a startup playground, but the ones who "make it" are those who can use their experience to play in the wider economy. Still, I feel like YC is providing a lot of false hope to people by not being more selective in the ideas they fund.

[omniscient_oce]

It's much lower hanging fruit than say developing an embedded hardware device that does something novel (and most likely more useful). Why not just aggregate some data about _other_ services and run machine learning.

[fireeyed]

Shovel for locating other shovels. Neat idea

[GKND95]

We've constantly asked ourselves this as we built our tech stack, love the idea!