HNHacker News
TopNewBestAskShowJobs

adam_gyroscope

893 karma · joined February 16, 2017

submissionscomments
adam_gyroscope··on Stripe backs $40M investment in AI accounting service Pilot.com
Please do ping me! My company was pretty starightforward, but I know they handle some fairly complex companies.
adam_gyroscope··on Stripe backs $40M investment in AI accounting service Pilot.com
We used pilot at my company (Gyroscope; sold to BlueVoyant in March last year). I had been doing my own books, which was annoying and not a good use of my time or something I’m good at. Pilot saved me a few hours a week for not too much money, totally worth it. (Full disclosure I know the founding team very well)
adam_gyroscope··on Prototool – A Swiss Army Knife for Protocol Buffers
We built https://github.com/GyroscopeHQ/grpcat at my company, which takes text-format protos as input and sends them to a gRPC endpoint. Looking at Prototool I think I should just merge the functionality into Prototool. This is cool!
adam_gyroscope··on Salsify – A New Architecture for Real-time Internet Video
This is great. I used to work on on both the traffic/network side of YT and the encoding side, and could see the value of this approach handled by distribution nodes at the edge of a network.
adam_gyroscope··on Pilot.com raises $15M to bring bookkeeping into the modern era
Oh, and I get a report every month with all my reports (balance sheet, P&L, etc) and summary data, and they do amazing customer support.
adam_gyroscope··on Pilot.com raises $15M to bring bookkeeping into the modern era
Yep - all of that is automated. Importantly, all the the classification of the transactions is automated as well, so I no longer have to look at a charge like "GUS99PWR-T BILL.COM" on my credit card statement and figure out a) where it is from and b) that it should be categorized as a marketing/promotional expense.
adam_gyroscope··on Pilot.com raises $15M to bring bookkeeping into the modern era
Pilot customer here; they are amazing. It's pretty different from a regular bookkeeper largely in that Pilot is super accurate and timely. My time spent doing bookkeeping for my small startup went from an hour or two a week to zero hours a week - those are pretty important hours. Highly recommended.
adam_gyroscope··on A computer science study plan
In 2012 I wondered if it was possible to build a 4 year CS degree from what Coursera offered at the time: http://www.thesimplelogic.com/2012/09/24/you-say-you-want-an... (spoiler alert: kinda, but not really). I bet you could easily do so now given the wealth of MOOC offerings.
adam_gyroscope··on Show HN: Pilot: Bookkeeping for startups, using humans and software
I'm a pilot customer; my company is an early (seed) stage software startup, and I use to do my bookkeeping by hand... and I had no idea what I was doing. "dubious thoroughness" is a great way to express my bookkeeping.

Pilot basically solved that problem. I just do my normal business and their integrations with my banks, expensify, credit cards, etc handle all of the transaction classification, and at the end of the month I get all of the statements I need to run a business. If there's any questions they reach out and help me address them.

I went from basically 4 founder-hours a month to zero, and I know my books are good. Totally worth it.

adam_gyroscope··on Stanford Compilers Course
I took this course a few years back when it was on Coursera and had taken a similar course many years ago in university; I've been a developer of some kind for 20+ years. This course brought together a bunch of stuff I understood but hadn't internalized so didn't use in my day-to-day job.

If you've been programming for a few years and take this course, you're going to say "OHHH that explains <some weird thing that happened in your career>" a bunch.

adam_gyroscope··on Firebase Predictions Beta
Great to see Google entering this space!

My company, Gyroscope Software (getgyroscope.com) does prediction, cohorting and timing of callbacks into mobile games and apps, with plugins for iOS and Unity. We've built a real-time prediction engine to make this happen; we also use some neat tech to get us enough data from the clients to do fast machine learning without the developer having to do any instrumentation, data collection, or other data wrangling - we do this automatically. Happy to answer any questions anyone has.

adam_gyroscope··on Ask HN: Who is hiring? (June 2017)
Gyroscope | Data Science | San Francisco, CA | REMOTE or ONSITE | Contract | https://gyroscope.cc

Gyroscope Software (https://gyroscope.cc) provides simple-to-use developer tools for maximizing mobile-user engagement via ML. Our long-term goal is provide developers with ML-based tools that give their software the capability to learn about users and act on that knowledge without requiring a PhD to build and understand.

Gyroscope uses cutting-edge techniques on a rich set of clean, homogenous data. We're creating a fast-iteration, highly-flexible data processing and analysis environment. Data scientists work hand-in-hand with engineers. Everyone writes code. We're built on Go, Python, R, Objective-C, & various cloud bits.

Feel free to email me at adam@gyroscope.cc if you're interested in working with us!

adam_gyroscope··on Air carriers are still failing people with disabilities
In some airports, airlines are essentially leasing gate space from the airport and any software the airline wants installed on the gate computer has to work in a very fixed set of constraints.

The shared gate agent computers are built from an industry-standard OS image (8 years or so ago this was Windows XP SP2, and not SP3, never SP3).

adam_gyroscope··on Air carriers are still failing people with disabilities
The six-character confirmation code you get when you book? That was the block address on disk of the start of your PNR.
adam_gyroscope··on Air carriers are still failing people with disabilities
Disclaimer: I built airline reservations systems with ITA Software.

I don't recall there being a minimum character count for SSR codes, but in legacy reservations systems (and the passenger name records (PNR) that contain these SSRs) there's for sure a max. Airline reservations systems never met a fixed-width format they didn't like, and the reason is that for a long, long time the unit of storage for a PNR was a block on disk and disk space was precious, so every few bits (literally bits, not bytes) counted. Those constraints have persisted through the ages.

Upthread someone wondered about disrupting this industry. It is expensive (there's a lot of moving parts) but it's basically engineering and integration, with the integration being the far harder part. For example, to participate in the airline networks (the networks that send messages to other airlines to handle cross-carrier booking, talking to GDSes, control departure, speak to DHS, etc) you need to understand the meaning of various airline protocol messages. These messages have unclear semantics; one airline may regard message A as meaning "create a ticket for this passenger" while another airline may regard the same message as "create a reservation for this passenger" which are two different things. The only way to know what these messages mean is to get on the phone with each airline and ask. The documentation is wrong and outdated; the meaning of those messages is the outcome they have to that specific airline and not anything else. Oh and the network connection you send these messages over is a VPN that you need to agree to setup with the carrier you want to talk to. Airline messaging is the reason most modern reservations system can only handle small airlines, since getting the semantics, routing, and other details of airline messaging requires too many humans and too many business relationships.

Additionally, switching reservations system can kill an airline. Most airlines operate with a day or less of cash reserves so if the switch to a new res systems stops operation of the airline for any amount of time you could kill the airline. There's no incentive big enough to switch to a new res system since the rewards are not immediately revenue-impacting.

adam_gyroscope··on Has Uber killed off its self-driving trucks?
Yep! Specifically: https://thelastdriverlicenseholder.files.wordpress.com/2017/... (that's them parked in the lot on Harrison) and then you'd erase all the OTTO branding :)
adam_gyroscope··on Has Uber killed off its self-driving trucks?
I walk past the Uber/Otto trucks every day (three or four are kept on Harrison St in San Francisco between 3rd and 4th), and while the trucks are still their with their very visible lidar/cameras/etc, in the last week they've removed all mentions of Otto from the trucks (previously they had the Otto branding on the doors).
adam_gyroscope··on Ask HN: Who is hiring? (May 2017)
Gyroscope | Software Engineer & Data Science | San Francisco, CA | ONSITE | Full-time | https://gyroscope.cc

Gyroscope Software (https://gyroscope.cc) is providing simple to use developer tools for optimizing mobile user engagement on a per-user basis using ML. We've just closed our seed round and hiring our first engineer. Our long-term goal is provide ML-based developer tools that solve real problems.

We're build on Go, Python, R, Objective-C, & various cloud bits. A detailed description of both jobs is at https://gyroscope.cc/jobs

Feel free to mail me at adam@gyroscope.cc if you've questions or are interested in applying!

adam_gyroscope··on Bringing VR Analytics to the Browser
Interesting point about the maturity & availability of 3rd party Unity libraries. Do you have any plans to open source your Unity libraries (UUID, etc)?
adam_gyroscope··on How the Airlines Became Cartels
If you are interested in the history of the airline industry up through the 90s, you should read Hard Landing. It's an excellent overview of how we got to where we are in the United States, and the meat of the book is about deregulation.

Note that "deregulation" in the airline industry (primarily) means the government no longer controlled which airlines flew which routes. Before deregulation, route ownership was decided by an oligopoly that would and could set high-margin prices on routes since the control of those routes was a) assured by the government and b) agreed on by all the major carriers.

After deregulation, airlines competed primarily on price. While this lead to the hub-and-spoke system and the low-fare carriers we see today, it also meant that almost anyone could fly anywhere in the US and abroad for (relatively) little money and a high amount of safety.

As an aside, airline reservations technology benefited immensely from deregulation, and the carriers that controlled the reservations systems ended up killing of the carriers that didn't (see, for example, Eastern Airlines). If you're interested in the tech of the airline industry, I recently gave a talk about this at Systems We Love: https://systemswe.love/archive/san-francisco-2016/life-of-an...

adam_gyroscope··on Life of a seed round: practical advice for fundraising
(author of the post, glad to see it is on HN!)

1. I'd do more research into each firm and partner. I think we could have been more tactical about who we talked to; in some cases it was obvious we didn't align with the investing thesis of the firm but this was our first time and we were so excited that anyone would talk to us that we took meetings that probably weren't as valuable to the VCs as they were to us. Also there's a million things in hindsight. Also this whole process is emotionally exhausting, so I'd prepare myself more for that aspect.

2. We're building our business - launching the product, selling it, getting to clearly defined revenue goals. The big thing we're doing now is making our key indicators obvious and tracking the heck out of them. We're really, really trying to stay focussed.

← PreviousPage 3 of 3