News for hackers. Stuff that matters.

kashishg

Thanks! Here's some more info on the git integration if you're curious: https://hightouch.io/blog/announcing-git-sync-back-your-sync...
Then you could say the same thing about Hightouch and Segment. Segment == Rudderstack for all intensive purposes.

The way it's differentiated is that Rudderstack does event collection + forwarding + some flavor of reverse ETL. In order to be successful with them you'd need to replace your whole stack and do event collection with them. A lot of companies already have an existing stack and they want to buy a best-in-class Reverse ETL player (something that takes 5 min to set up) and that's where we come in!

Thank you for the kind words :) We are just getting started!
we started building this one August 2020 but honestly just had a lot of fun working on the design and UX! Conveying your feedback to him now!
On a high level, ETL/ELT is about sending data from your SaaS tools into your data warehouse (you are reading from different tools). Reverse ETL is about getting data from your warehouse into tools (writing into different tools). Building ELT is a fundamentally different technical challenge than building Reverse ETL. Aspects like types, rate limits, and destination state (knowing whether data already exists in a destination) are unique to Reverse ETL. Visibility becomes challenging too as some destinations have unique quirks, like API contracts where you write to them but you don’t know if the write was successful or completed until later. Writing to tools also requires references between objects (foreign keys onto existing data, like mapping Companies and Opportunities in Salesforce) that aren’t necessary in the ELT world.

From a product perspective, the UX is very different as well. Reverse ETL requires a lot more user input (ex: mapping which fields to update in a tool), whereas ELT typically mirrors data using a standard schema (without much user customization involved).

We are close partners with ELT tools like Fivetran, and you can see our partnership post here: https://fivetran.com/blog/fivetran-partners-with-hightouch-t...

Great question! We have a whole page about this here (https://hightouch.io/blog/hightouch-vs-census/).

But the TLDR is that Hightouch has more developer focused features (like a live debugger, alerting, version control with Git, and more here: https://hightouch.io/data-features/), a dedicated UI for business users to visually filter models (called Hightouch Audiences), more transparent pricing, as well as more integrations (70+) that are also deeper and customized for each tool.

Yes, Clickhouse is a top priority source for us to build next to enable real-time analytics (we already support Rockset as a source used by customers like Seesaw). Would love to learn more about your use case for Clickhouse: feel free to reach out to [email protected]!
Thanks so much for your support from the very beginning! We've only been in market publicly for a little over a year now actually. We figured better late than never (and it still feels early for us!) :)

Launch HN: Hightouch (YC S19) – Sync data from data warehouses to SaaS tools

Hey HN! Kashish, Tejas, and Josh here. We’re building Hightouch (https://www.hightouch.io/), a reverse ETL platform— that is, software that gets your data back out of your data warehouse and into the SaaS tools that people at your company are familiar with (like Salesforce). We enable you to Bring Your Own (BYO) database so that all your SaaS tools run off of the same dataset. You specify what...

( Read More... | 47 comments | 132 points | on HN )
Re:Too Many SaaS Tools: Turn Your CRM into a PLG CRM on Thursday August 19, 2021 @03:24PM (#28238174)
We would love feedback on our approach and to learn how you are doing this at your company!
We're partnered with a travel agency with GDS access. Additionally, we use various other OTAs that give us the full range of inventory for hotels and flights.

In general, travel agents that only use GDS don't have access to budget airlines such as Easyjet or alternative accommodations such as Airbnb or Sonder.

I wonder why enterprise tools end up this way, is it because the buyer (CFO/procurement) is different from the end user (employees)?
We too are hopeful that the narrower scope will allow our economics to work out!
Recently discovered that Google flights not only has a matrix when you're choosing dates, but also has an even better one after you've already chosen dates to view the price differences for +/- 5 days on both the departure and arrival.

Just click on "Date Grid" after you search for your query and it will give you a matrix. I can't believe I just found this!

This one is the best by far, Google Flights 2nd place. Sometimes our agents even use the Skyscanner matrix for our users because it's so good.
We've heard a lot about flightfox! We're focused employee-first so that we can build a tool that employees their managers for, but that being said, several of our corporate customers have been seeing a similar cost savings because of the ability to set travel policies with Carry.

Just curious, what % of travel at your company is booked through flightfox?

Yes! We're working on the calendar integration :) We find that several business users will omit calendar because of privacy or because they don't keep them up to date, but power users of GCal would love this I think
We've been playing around with the SaaS idea that you described, because there are 80k+ travel agents in the US and they are underutilized! We eventually decided that performing the travel booking directly is important because that way we control the user experience, and our level of service is really the main value prop right now

We'd actually love to talk to your corporate travel provider and see if a partnership makes sense! ping me at kashish [at] carry [dot] travel? Otherwise we can get you started as an individual if your company allows it :)

You're right, flights/travel is so mission critical that we don't see "AI" taking over this workflow anytime soon. There will be a real person verifying every decision for the near and medium-term future. The automation component helps these real-people make decisions faster (ideal case is 30 seconds or less once it gets really good)
Thanks Adhityaa! So cool that you found us on here!

You're so right, there are no solutions that focus on point A -> B door-to-door, and we've found that the reason for that is that most solutions don't know their end customer. They don't include information like home address and office addresses. The only solution that we can see handling end-to-end is Google Flights since they can easily integrate with Google Maps, for us, a very important goal is to help people manage their day of travel, not just their bookings :)