siddharthb_

5 ms·

1.
Edit Videos with Meta AI(about.fb.com)
3 points by siddharthb_ 1y ago|0 comments
2.
by siddharthb_ 2y ago
TurboML ( https://turboml.com ) | Real-Time AI platform | Remote AI Engineer: * Build and curate high-quality datasets, leveraging them to train, fine-tune, and optimize Large Language Models. * Responsibilities include data pipel
3.
Fluss is now Open Source(ververica.com)
1 points by siddharthb_ 2y ago|0 comments
4.
by siddharthb_ 2y ago
TurboML ( https://turboml.com/ ) | Remote | Software Engineer / ML Engineer TurboML is an ML platform reinvented for real-time. We specialize in bringing real-time ML—similar to the tech behind TikTok—to improve use case
5.
Progress Acquires ShareFile for $875M(techcrunch.com)
1 points by siddharthb_ 2y ago|0 comments
6.
by siddharthb_ 2y ago
TurboML ( https://turboml.com/ ) | Remote | Software Engineer / ML Engineer / Enterprise Sales / Technical Writer TurboML is an ML platform reinvented for real-time. ML is data-driven, and TurboML turbo-boosts
7.
by siddharthb_ 2y ago
TurboML ( https://turboml.com/ ) | India | Enterprise Sales Manager/ Software Engineer / ML Engineer TurboML is an ML platform reinvented for real-time. ML is data-driven, and TurboML turbo-boosts your ML with real-
8.
by siddharthb_ 3y ago
TurboML ( https://turboml.com/ ) | India | DevOps / MLOps / ML / Streaming Systems / UI Designer TurboML is an ML platform reinvented for real-time. Imagine having all the capabilities of an ML platform –
9.
by siddharthb_ 3y ago
TurboML ( https://turboml.com/ ) | India | Remote | Full-time/Intern | Business Development Manager / Technical Content Architect TurboML is a real-time machine learning platform designed for fast-paced ML use cases
10.
Databricks acquiring AI startup Einblick(finance.yahoo.com)
3 points by siddharthb_ 3y ago|0 comments
11.
Flow Acquired by Deloitte(betakit.com)
1 points by siddharthb_ 3y ago|0 comments
12.
by siddharthb_ 3y ago
TurboML | India | Remote | Full Time / Intern | MLOps / DevOps / Streaming / Full Stack We're building a real-time ML platform for fast-moving use-cases where data freshness and low latency are of utmost importance.
13.
by siddharthb_ 3y ago
TurboML | Founding Engineer | Remote | Full Time | Real-Time ML We're building a real-time machine learning platform for fast-moving ML use-cases such as fraud detection, ad bidding, dynamic pricing, recommender systems. We offer conti
14.
Cisco Acquires Splunk(splunk.com)
923 points by siddharthb_ 3y ago|502 comments
15.
What is the cost of being outdated for you?(twitter.com)
1 points by siddharthb_ 3y ago|2 comments
16.
by siddharthb_ 3y ago
Just as any asset depreciates over time, so does data. Ask yourself how long it takes for you to update your models with today’s data? Is it weeks, or perhaps months? Imagine using stale gasoline in a Ferrari. It might chug along, but with
17.
Can Amazon Survive ONDC in India?(pib.gov.in)
1 points by siddharthb_ 5y ago|0 comments
18.
by siddharthb_ 5y ago
We design a classification algorithm called Clustering Aware Classification (CAC), to find clusters in data that are tailor made to be easily classifiable when used as training datasets by classifiers for each underlying subpopulation. CAC
19.
Show HN: Don't Just Divide; Polarize and Conquer(github.com)
9 points by siddharthb_ 5y ago|1 comments
20.
by siddharthb_ 5y ago
Existing streaming graph anomaly detection methods only detect edge or subgraph anomalies. We extend count-min sketch to higher-order preserving the dense subgraph structure & detect both. Our approach is the first streaming method that
21.
Show HN: Higher Order Sketch-Based Graph Anomaly Detection(github.com)
55 points by siddharthb_ 5y ago|1 comments
22.
by siddharthb_ 5y ago
Thanks! Great questions. Existing papers switched the class for KDDCUP99 so we followed them to have a fair comparison. We also use UNSW and DoS which are time-series based datasets. Once the encoder is trained on few normal samples and use
23.
by siddharthb_ 5y ago
MemStream detects anomalies from a multi-aspect data stream and outputs an anomaly score for each record. MemStream is a memory augmented feature extractor, allows for quick retraining, gives a theoretical bound on the memory size for effec
24.
Show HN: Memory-Based Anomaly Detection in Multi-Aspect Streams(github.com)
36 points by siddharthb_ 5y ago|3 comments
25.
by siddharthb_ 5y ago
The course is taught by both Yann LeCun and Alfredo Canziani. In fact, it is hosted on Alfredo's webpage. Can you please choose a suitable headline?