suprgeek
Joined 17 years ago
Karma
14,461
Submissions
1,136
Member since
Jul 1
Recent submissions
Comment · 5 months ago
CData | Forward Deployed Engineer + Principal FDE (Agentic Systems) | Remote | Full-time
Details of the Job here: https://ats.rippling.com/cdata-software/jobs/bd754ddd-30c4-4...
https://ats.rippling.com/cdata-software/jobs/e9d1d4fa-18c9-4...
Remote with possible some travel
Comment · 8 months ago
Since the recent discoveries about Shingles Vaccine delaying dementia
https://hsph.harvard.edu/news/link-between-shingles-vaccine-...
One of interesting checks in this study might be to check when (if) any of the participants had taken this vax and what the impact might be on an Alzimer's diagnosis.
Comment · 1 year ago
In the era of anything even remotely AI-adjacent raising ridiculous amount of money - and proving to vaporware - this is one raise that makes me happy.
We need more of the Bryan mindset (and less of Larry) to ensure that when the AI-Enabled future shows up we have some semblance of a fair society that works for all.
https://www.youtube.com/watch?v=-zRN7XLCRhc&t=2047s
Comment · 1 year ago
Don't be put off by the reported high-cost
Make it possible->Make it fast->Make it Cheap
the eternal cycle of software.
Make no mistake - we are on the verge of the next era of change.
Comment · 2 years ago
Great project and excellent initiative to learn about embeddings.
Two possible avenues to explore more.
Your system backend could be thought of as being composed of two parts:
|Icons->Embedder->|PGVector|->Retriever->Display Result|
1. In the embedder part trying out different embedding models and/or vector dimensions to explore if the Recall@K & Precision@K for your data set (icons) improves. Models make a surprising amount of difference to the quality of the results. Try the MTEB Leaderboard for ideas on which models to explore.
2. In the Information Retriever part you can try a couple of approaches: a.after you retrieve from PGVector see if you can use a reranker like Cohere to get better results https://cohere.com/blog/rerank
b.You could try a "fusion ranking" similar to the one you do but structured such that 50% of the weight is for a plain old keyword search in the metadata and 50% is for the embedding based search
Finally something more interesting to noodle on - what if the embeddings were based on the icon images and the model knew how to search for a textual descriptions in the latent space?
Comment · 3 years ago
A few items that did not make it into the list:
1) Adani is very close to the current ruling party and the PM - deploying his extensive business network and assets (in various forms) to help get the PM elected (2014) and then re-elected. A bit of googling will reveal the vast scale of politico-corporate nexus between these two - extending all the way back to when they were chums in Gujarat - the Indian state of the PM
2) The only reason the Adani Group has not been hauled up in-front of multiple regulators and law-enforcement agencies on a constant basis is because of #1 above
3)Adani ports seem to be curiously immune to investigations despite literal TONS of drug seizures [1]
4) There were multiple stories of forced resignations being used by the companies across its workforce to trim costs during the pandemic and avoid paying hefty severance per India's strict labor laws
5)A lot of the public sector banks in India are purported to be under pressure to not recall loans or issue fresh loans on demand to him - essentially a federal backstop to all his liabilities - enabling crazy levels of debt
[1] https://www.ndtv.com/india-news/adani-ports-says-its-termina...
Comment · 4 years ago
The AWS scaling series https://aws.amazon.com/blogs/startups/scaling-on-aws-part-1-... gives you a very nice primer on this exact situation.
You may or may not opt to use AWS - the fundamental concepts would translate well over other cloud providers or even on premises.
Then you can follow along parts 2, 3 &4 to scale up by factors of ~10 or more -
https://aws.amazon.com/blogs/startups/scaling-on-aws-part-2-... https://aws.amazon.com/blogs/startups/scaling-on-aws-part-3-...
Comment · 6 years ago
This was somewhat predictable.
I will be really interested in another piece of data.
When we re-open fully (whatever that means) and people get back on the roads perhaps even more than they used to will it reset the waiting times because of excess collisions and more fatalities?
Comment · 7 years ago
A lot of the suggestions in this post are "idempotent" only in the sense of the specific use case that the author is interested in. Please do not take this advise to be generally applicable for the uses of "idempotent" or in general to mean best practice while writing bash scripts.
The advice such as replace rm <FileName> with rm -f <FileName> could lead to Disaster depending on the scenario. So a HUGE YMMV
Comment · 7 years ago
Finally!
Now please also admit that the TouchBar is a colossal failure that only a few niche users grew to like and either:
1) Ditch it completely - return the Fn key row back but add button for Touch Id (a. la the new MacBook Air) OR
2) Invent a better version with Haptic Feedback so that Touch typists get a FEEL for hitting keys instead of moronically having to look down to ensure that they did not F* things up
AI/ML, Cloud computing, Java, Big Data, Startups, Health