Hacker News Reader

Avantos.ai | https://avantos.ai | Sr Software Engineer | St. Louis or NYC | 150-200k

We're building an AI notetaker and task management tool for financial advisors. Startup is headquartered in NYC. If you're not in NYC, quarterly travel to the office is expected.

Front end: React DB: Postgres APIs: Golang, Elysia/Node Infra: AWS ala Terraform

Interview Process

1. 10-minute async coding challenge 2. Hiring manager chat and 20-minute coding challenge OR take home challenge 3. 45-minute Coding Challenge w/ another Engineer from the team 4. System Design Interview 5. Chat w/ a Product Manager

Apply here: https://avantos.breezy.hr/p/4a707ef4f952-senior-fullstack-so...

kmdupreeon Ask HN: Who is hiring? (February 2026)

Avantos.ai | https://avantos.ai | Sr Software Engineer, AI Team | St. Louis or NYC | 150-200k

We're building an AI notetaker and task management tool for financial advisors. Startup is headquartered in NYC. If you're not in NYC, quarterly travel to the office is expected.

Front end: React DB: Postgres APIs: Golang, Elysia/Node Infra: AWS ala Terraform

Interview Process

1. 10-minute async coding challenge 2. Hiring manager chat and 20-minute coding challenge 3. 45-minute Coding Challenge w/ another Engineer from the AI team OR take home challenge. 4. System Design Interview 5. Chat w/ a Product Manager

Apply here: https://avantos.breezy.hr/p/4a707ef4f952-senior-software-eng...

kmdupreeon Ask HN: Who is hiring? (January 2026)

Avantos.ai | https://avantos.ai | Sr Software Engineer, AI Team | St. Louis or NYC | 150-200k

We're building an AI notetaker and task management tool for financial advisors. Startup is headquartered in NYC. If you're not in NYC, quarterly travel to the office is expected.

Front end: React DB: Postgres APIs: Golang, Elysia/Node Infra: AWS ala Terraform

Interview Process 1. 10-minute async coding challenge 2. Hiring manager chat and 20-minute coding challenge 3. 45-minute Coding Challenge w/ another Engineer from the AI team 4. System Design Interview 5. Chat w/ a Product Manager

Apply here: https://avantos.breezy.hr/p/4a707ef4f952-senior-software-eng...

kmdupreeon Ask HN: Who is hiring? (December 2025)

Avantos.ai | Senior Software Engineer, AI Team | NYC or St Louis | 150-200k

We're building an AI notetaker and task management tool for financial advisors. Startup is headquartered in NYC. If you're not in NYC, quarterly travel to the office is expected.

Front end: React DB: Postgres APIs: Golang, Elysia/Node Infra: AWS ala Terraform

Apply here:

https://docs.google.com/forms/d/e/1FAIpQLSd4JRq-lvpsA8GGnyrd...

kmdupreeon Ask HN: Who is hiring? (November 2025)

Avantos | Lead Fullstack Engineer | Remote (US) | 150-200k

We're building an AI notetaker and task management tool for financial advisors. Company is headquartered in NYC. Although the role is remote, quarterly travel to the office is expected.

Front end: React

DB: Postgres

APIs: Golang

Infra: AWS ala Terraform

Apply here:

https://docs.google.com/forms/d/e/1FAIpQLSd4JRq-lvpsA8GGnyrd...

kmdupreeon Ask HN: Who is hiring? (May 2025)

Avantos.ai | Jr Software Eng | Remote (UTC-6 - UTC+2) | 60-80k

Avantos is an AI platform that streamlines client onboarding and servicing for financial institutions. Its a Series A startup headquartered in NYC.

Tech Stack Languages: Typescript, Golang Frameworks/Libs: React, Next.js, Vitest, Playwright DB: Postgres Infra: AWS, Terraform, Docker

React experience is a must have. Experience with the rest of the stack is preferred, but not required.

You'll get a sense of what we're building during the coding challenge, which you can start by sending a POST request to: https://apply-to-avantos.dev-sandbox.workload.avantos-ai.net Payload should be this: {"email": "<your-email>"}

You'll need to make sure the user agent isn't something too bot-seeming or WAF will block you. This user agent works: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36

kmdupreeon Ask HN: What are you working on? (April 2025)

http://makefizz.buzz/

I work at a small startup trying to hire engineers and got tired of looking through resumes. As Joel Spolsky points out, they're not great indications of technical ability. Instead, I decided to throw together a take-home challenge that applicants could access via an API. "If they can't solve the challenge, I don't need to see their resume," I told myself.

FizzBuzz.md is a better version of the solution I built for work. It lets applicants send questions and submissions to configurable email addresses via API so the email addresses aren't exposed directly. (Less spam, FTW)

kmdupreeon Ask HN: Who is hiring? (March 2025)

Avantos.ai | Frontend Eng | Full-time | Remote (USA)

Series A AI Fintech Startup headquartered in NYC. If you’re outside of NYC, we'd like you to make quarterly visits and are happy to support higher-frequency visits.

You'll get a sense of what we're building during the coding challenge, which you can start by sending a POST request to:

https://apply-to-avantos.dev-sandbox.workload.avantos-ai.net

Payload should be this:

{"email": "<your-email>"}

You'll need to make sure the user agent isn't something too bot-seeming or WAF will block you. (Sorry we're a startup no time to fix.) This user agent works: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36

kmdupreeon Swift Homomorphic Encryption

it says on their webpage that they aren't using HE

kmdupreeon Internal tools often make bad startup ideas

OP here. Yes! Thank you for seeing this! I make this point explicitly in a previous post called "Conflicting employee vs. business incentives slow B2B SaaS growth rates"

kmdupreeon Internal tools often make bad startup ideas

iirc, dropbox was actually very technically challenging, which is why they tested the idea with a mock landing page first before building it (source: Lean Startup) and steve jobs famously spoke with drew about what he'd built early on

kmdupreeon Internal tools often make bad startup ideas

OP here with a friendly reminder that we should also look at the denominator when evaluating how useful the "internal tools make good startups" heuristic is.

Also note that I'm not claiming they are never good ideas, so pointing out instances where they've panned out isn't very interesting. It'd be more interesting if you could show that the hit rate among these types of startups is better than I'm suggesting (for that we need to consider the denominator).

kmdupreeon Internal tools often make bad startup ideas

OP here. very interesting point!

kmdupreeon Figma and Adobe abandon proposed merger

Probably better for the consumer.

kmdupreeon The UIs ChatGPT Won't Replace

OP here. Thanks for bringing up memory palaces! They're a great example of how important our sense of space is for understanding anything!

kmdupreeon The code that ChatGPT can't write

Hey, there! OP here. This would be a neat experiment to try.

One related experiment, however, already suggests the result. Someone tried to get ChatGPT to solve advent of code challenges: https://github.com/golergka/advent-of-code-2022-with-chat-gp.... These challenges are very clear, and it already seems to struggle to get the answers. One the second day, it took 12 tries to get it right. If it struggles this much with clear requirements, I don't think it'll do well with vague ones.

kmdupreeon Show HN: A programmable Jupyter notebook assistant for VSCode users

I'm the guy working on data chimp. Here's a quick overview:

data chimp automatically shows data visualizations, tables, and messages about your data as you work in your Jupyter notebook according to rules you set. The rules are specified in code via special “config” notebooks, and when the rules are triggered, your rule code has access to the data frame that’s currently being analyzed, so you can visualize or aggregate it however you like.

For example, you can write a code rule that says, “if I’m working with a data frame column that has more than 3% missing values, show a time series plot of the percentage of missing values for the column over time.” With rules like this, you can spot unexpected features in your data, catch buggy data wrangling code, get oriented in a new data set quickly, or encourage analysis best practices within your team.

I built data chimp because when I first joined the data science team at Heap (I came from an engineering team), I was shocked at how error-prone and repetitive exploratory data analysis was. On one occasion, we published an analysis that contained a bug that slipped past code review, and when our stakeholders got excited about the analysis, we had to tell them to hold off while we re-crunched it. Moreover, I was regularly frustrated by being forced to either

* write the same visualization and aggregation code over and over again

* use a canned function (e.g., pandas profiling’s ProfileReport) that wasn’t flexible enough to show me what I needed to see in a particular scenario and that stifled my ability to iterate on the visualizations generated by the function

data chimp makes analysis less error-prone by borrowing some ideas from automated testing and linting, and it resolves the repetitive code dilemma by making it easy to customize what is automatically shown according to your own rules and by making it possible to paste the code that generated a particular result back into a notebook cell so you can easily iterate on it.

data chimp is currently “beta quality,” but I’m hoping to get some feedback by posting here, and more generally, I’m looking to understand the problems that data scientists tend to face as they’re working with data. If you want to tell me why data chimp sucks or complain about your job to me for a few minutes, I’d love to chat sometime. ;) Book a time here: https://meetings.hubspot.com/matt-dupree

kmdupreeon Postgres audit tables saved us from taking down production

RETURNING * is a clever way to do this. What is this * from affected_rows syntax you're using here? Any chance you have a link to the docs that describes this?

kmdupreeon Postgres audit tables saved us from taking down production

clever! hadn't thought of combining this with hstore.

kmdupreeon Postgres audit tables saved us from taking down production

Thanks for sharing these links!

kmdupreeon Postgres audit tables saved us from taking down production

Ah. Sorry to disappoint. The technique was just new to me, so I thought it was worth sharing. :)

kmdupreeon Postgres audit tables saved us from taking down production

Author here. :)

>I have used it in at least one application before.

Good to know that other folks are using this technique!

>One of my favorite things about it is that you can use it even when you don't own the application writing to the database

Great point! Hadn't thought of this.

kmdupreeon Rolling out 1M Postgres indexes

Great question! I should have been clearer. Each event type gets a new partial index where the WHERE clause of the partial index corresponds to the properties of the new event type.

Hope that clarifies! Lmk if not. :)

kmdupreeon The Million Indexes Incident

Great question! For new events we have 8 workers that process multiple customers at a time so the time to querability is typically pretty fast!

kmdupreeon Working around a case where the Postgres planner is “not very smart”

Thanks for reading!

I hear you on visibility maps being intimidating. In practice, I haven't seen any cases where visibility map issues have prevented an index-only scan. But we did initially think that the visibility map was to blame for what we were seeing!

RE space of the index: it cost us about 1% of the free disk space on our workers. It was worth it for this particular feature.

kmdupreeon Working around a case where the Postgres planner is “not very smart”

You're exactly right that our schema could be better and that it's non-trivial to execute on a schema change, especially because we actually run a distributed Postgres cluster via Citus AND we use a special sharding method that we manage manually.

We actually just started working towards how we might do a schema change for the 1+ million shards in our cluster. Hopefully, we'll be able to write up some learnings on the schema change after its done. :)

kmdupreeon Building Better Code Review at Heap

Glad we agree! (I'm the author)

Less bike shedding => more substantial review comments => better code

kmdupreeon Show HN: InStock – like Waze, but for grocery availability

We've updated the app so that it searches more broadly (10 store instead of 5), let me know if that helps! :)

kmdupreeon Show HN: InStock – like Waze, but for grocery availability

Appreciate the feedback! Posted it in our slack channel now. :)