775 karma · joined October 1, 2024
Most of what people call "data science" is now doable with an LLM and a weekend. I focus on the messy, ambiguous, regulation-heavy gnarly problems where nobody is sure what the right question is. With AI as an execution lever, of course.
What I've done:
- Saved a German automaker from lemon law recalls. ($20M+ exposure avoided)
- Real-time emissions detection for industrial smokestacks. ($75K fines per incident prevented)
- Revenue optimization and persona identification for debt collections (15-20% net revenue lift)
- LLM-based legal document extraction, reducing due diligence turnaround from 3 days to 30 minutes
- Built vessel piracy risk engine that informed naval escort deployment, contributing to fewer hijackings
Things I'm unwilling to work on: - Gambling.
- Ads/Surveillance.
- Payday loans/rent-to-own.
Email me with a one-paragraph description of your problem. If it's weird enough, I'll reply within 24 hours.Maybe a consultancy of people who have seen and done it all before - Very selective of their clients.
But the money was so good we (The royal we) didn't think we needed it, that would just get in the way. Did you see how much FB employees were getting paid in 2015! Insanity! Now, even the skutters have a better union than us.
A plumber, or an electrician has a better union, and hence rights and protection than us.
But if you're building a brand new field you can still build a guild.
Which is the thing, we do need a single key that can be used for all those things. So we get PGP.
And in this case smaller is better?
Now, I'm master of about a thousand lines of pricing code plus documentation and research which actually matters. The AI can handle the rest as a very skilled junior with a TBI.
You absolutely are a proctor, or senior manager. The AI is the smartest most well read junior you will ever meet, but don't go out of its happy path.
As you go out of the commonly read happy path for CRUD apps, you'll have to get more and more involved. I wouldn't write a new kernel design with AI right now, I might write a Linux kernel driver with it though.
They already are. I’m successfully using frameworks like bmad to deliver complex apps at that level. My job is to manager the see, as, ux, sre processes and catch errors.
I spend more time refinding prd , epics and stories than I do elbows deep in code.
If I don’t like the output of a story I nuke it change the story and have the flanker try again. I’m using the open source glm, kimi, deepseek models. I expect the full pipeline to be good enough by the end of the year.
They built an entire wafer ASIC. The entire thing is one huge active ASIC. it takes a lot of cool engineering and cooling to make it work, and is very cool.
NIN / The Fragile
Radiohead / OK Computer
Daft Punk / Homewerk
All of these are generation defining albums.
I'm a data scientist with over 20 years of experience specializing in solving hair on fire problems. I thrive on gnarly problems AI can't complete even if walked through the problem step-by-step. Often it involves getting hands on or talking with staff to find the data that's really needed.
My past work includes:
- Saving a German automaker from lemon law recalls.
- Helping a major cloud vendor predict server failures to enable load shedding.
- Real-time on demand routing logistics work .
- Airline flight delay forecasting.
- Oil & Gas forecasting.
- Shipping piracy risk.
- Wound identification and classification.
- Revenue optimization, persona identification and dynamic "risk-on/risk-off" risk management for ARM.
I'm currently working on spec work with criminology typologies of victim disposal in serious crime. The tigger for this is the Justo Smoker murder of Linda Stoltzfoos, Police spent months searching in the wrong places, when a good behavioural model would have shortened the search time signifcantly. I'm unlikely to get it in the hands of law enforcement, but I like solving a good problem.Things I'm unwilling to work on:
- Gambling.
- Ads/Surveillance.
- Payday loans/rent-to-own.
Get in touch if you have a really difficult problem you're trying to solve. Email in profile.Will new developers know/understand what they don't know, or will the new state of things simply become normalized?
Will the SEC decide to?
It's about a company that makes sausage that discovers their meat supply is full of sawdust and rat feces. They decide to sell off all their inventory, and crash the market, before their competitors do.
https://512pixels.net/2024/03/apple-jonathan-modular-concept...
Is HN run on eventual consistency? My numbers don't match yours.
I'm a data scientist with over 20 years of experience specializing in solving hair on fire problems. I thrive on gnarly problems AI can't complete even if walked through the problem step-by-step. Often it involves getting hands on or talking with staff to find the data that's really needed.
My past work includes:
- Saving a German automaker from lemon law recalls.
- Helping a major cloud vendor predict server failures to enable load shedding.
- Real-time on demand routing logistics work .
- Airline flight delay forecasting.
- Oil & Gas forecasting.
- Shipping piracy risk.
- Wound identification and classification.
- Revenue optimization, persona identification and dynamic "risk-on/risk-off" risk management for ARM.
I'm currently working on (semi) automating the DS loop, building workflows on top of BMAD method AI workflows. I say semi, because some problems you simply must get out of the office and speak with people or find data the AI doesn't have access to. Many problems will have e2e solves in an automated fashion.Things I'm unwilling to work on:
- Gambling.
- Ads/Surveillance.
- Payday loans/rent-to-own.
Get in touch if you have a really difficult problem you're trying to solve. Email in profile.I'm a data scientist with over 20 years of experience specializing in consulting and fractional leadership. I do the data science that AI's can't do. I thrive on gnarly problems where standard off-the-shelf solutions fall short, and recently where Ai just can't complete the task even if walked through the problem. My track record includes saving a German automaker from lemon law recalls and helping a major cloud vendor predict server failures to enable load shedding.
I've tackled a wide range of challenges across various industries, including oil reservoir and well engineering forecasting, automotive part failure prediction, and shipping piracy risk to route ships away from danger. My technical work extends to realtime routing (CVRP-PD-TW) for on-demand delivery, legal entity and contract term extraction, and wound identification with tissue classification. I also work with the current wave of LLMs and agents, and make them do magic.
I've worked with the standard stacks you’d expect: Python, PyTorch, Spark/Ray, AWS, Agentic engineering, etc. But I believe the solution must be driven by the problem, not the tools. I bring years of experience helping companies plan, prototype, and productionize sane data science solutions.
Please reach out if you have a difficult problem to solve. I do love stuff in physical meat-space.
NB: Please do not contact me if you are working on ads, gambling, or "enshittification". I prefer to sleep at night.
Keep the interpreters surface area as small as possible. Do some work to make sure you can accelerate numeric, and JIT/compile functions down to something as close to native as you can.
Wolfram, and Taliesin Beynon have both said Wolfram were working internally to get a JIT working in the interpreter loop. Keep the core small, and do that now while it's easy.
Also, it's just easier to write in Mathematica. It's probably 10x smaller than the rust code:
f[x_Integer]:=13*x;
f::help:="Multiplies x by 13, in case you needed an easy function for that."
EDIT: Another important thing to note is the people who really deeply know specific subjects in math won't be the best, or even good rust programmers. So letting them program in woxilang will give the an opportunity to contribute which they wouldn't have had otherwise.There is an architectural flaw in Woxi that will sink it hard. Looking through the codebase things like polynomials are implemented in the rust code, not in woxilang. This will kill you long term.
The right approach is to have a tiny core interpreter, maybe go to JIT at some point if you can figure that out. Then implement all the functionality in woxilang itself. That means addition and subtraction, calculus, etc are term rewriting rules written in woxilang, not rust code.
This frees you up in the interpreter. Any improvements you make there will immediately show up over the entire language. It's also a better language to implement symbolic math in than rust.
It also means contributors only need to know one language: woxilang. No need to split between rust and woxilang.