mh44
mh44 on Hacker News
Posted by
mh44 on Thursday February 04, 2021 @09:55PM
from the ask-hn dept.
I'm struggling with thinking about operations decisions: "mark this account as fraud" or "email customer if they have an item in their cart for >3 days". Decisions that define the customer experience or business operations of your company.I've seen a combination of manual, expert rules/code and machine learning involved at my company, but it's a mess. So many teams, so much politics. How...
We looked into these, but they honestly seemed like a pain to integrate + getting ops to learn a new DSL would need org changes I can't really influence. Have you used these? Would love to hear more! I'd much rather not build it myself haha
Thanks! This feels like a Zapier/IFTT? What I've found with those tools is that they help me set up the connections, but I often find myself still unable to pass the "data" responsibility over to the business user.
What are your most commonly supported use case examples?
Thanks for sharing your story. Oh man, that sucks! Yeah, it's funny there's things I don't want to build in the first place. And then the moment you build them, you are responsible for making sure they work for the rest of your tenure. Managing downstream dependencies is genuinely hard for so many systems, and is made worse by the lack of visibility and custom code-writing for internal tools.
Makes sense. Thanks for the tip on linking Excel!
Wish you'd made this two months ago - had to build a lot of this from scratch at my startup and would've totally used something like this! I tried looking on the internet, and found basically nothing.
Oh interesting, will look into that! Excel works for one-offs, but with repeated queries, or if they're exploring data that they don't know exactly how they'd use, or if they want to affect changes in a system ("send email to all these customer ids"), they'd still have to come back to me?
Makes sense, that's what I've seen other folks do. I'm curious if you've seen them ask for more features over time, more actions, ability to manage the rules they write - and whether the full blown version of what they need was easy to support?
Posted by
mh44 on Friday January 29, 2021 @01:33PM
from the ask-hn dept.
I work with operations teams that need to frequently update very simple business logic (think SQL). Other engineers who work with operations teams (customer success, fraud/risk, content moderation), do you use internal tools to make you productive? Or enable ops to self-serve via low-code?Looking to spend less time updating logic and more time building systems.
something I wish existed. Operations teams have business context, but no direct control over even simple workflows. Engineers are overwhelmed updating rules logic, there’s no visibility/auditability and customer experience tanks. I’m building a tool to combine the best of humans, rules and ML decisions - by writing SQL to trigger actions. e.g. “if transaction > $1M and account < 2 days old, block and review” Best explained by a demo
https://www.loom.com/share/57006a1065244589beec80fb890b0dc8
Looking for beta users! If you’ve felt this, would love to chat.
This. Nothing wrong with looking at the bright side, but like with everything, there’s a balance. When confiding in Pollyanna friends, you walk away either feeling not understood, or worse, thinking they’re liars, even to themselves.
The most important consideration here is whether you want to take VC funding at all, or bootstrap as much as possible. There are many fine businesses that fall into the VC game, have undue pressure to hit hockey stick growth and hire too fast. If their strategic prospects are limited, you end up having to downsize and restart.
This essay changed my perspective
https://sahillavingia.com/reflecting Not saying this model is right for you, but worth a consideration.
Please tell me you’re at least if donating the $ to charity!