Mayank Shrivastava

Learner | Builder | Leader
Apache Pinot PMC & Committer

[Mayank Shrivastava @mayankshriv]

So much to learn, so little time on this rock.

[StarTree @startreedata]

🏇 The Kentucky Derby is over in 2 minutes—so why wait hours for analytics?

⚡️Apache Pinot🍷 delivers real-time speed that traditional warehouses can't match. For fast decisions, bet on Pinot.

Before you place your bets! Hear from @FetchRewards below!🔗
#RealTimeAnalytics #data

[Kaivalya Apte - The Geek Narrator @thegeeknarrator]

It is so cool to see Apache Pinot winning 2024 BigDATAwire editor’s choice award.

I learned about Apache #Pinot back in 2021-22 when I was trying to figure out a better way to do realtime analytics. I stumbled upon so many great articles from Startree and some youtube videos that helped me get started.

Most importantly the community is filled with amazing people who not only bring an extensive experience working with high scale distributed systems but also have an attitude to go extra mile in helping anyone who is new and getting started.

I remember I faced many questions and challenges mainly because of my lack of understanding, but folks like @mayankshriv @KishoreBytes @nehapawar18 @richardstartin helped me personally.

Great people, make great communities that makes an open source project massively successful like Apache Pinot.

Another important point that is worth noting is that the code is pretty readable, well maintained and structured. I remember Neha Pawar shared how the team was able to implement tiered storage very quickly (more like a hackatho?) mainly because the abstractions are clean.

So don’t hesitate and try out Apache Pinot and join the community.

Here you can read more about the award:

https://t.co/7AsIg9Fo68

Some episodes I did related to Pinot can he found below:

Tiered Storage implementation by StarTree (Apache Pinot) with Neha Pawar
https://t.co/aO68D3M0DI

Running Distributed Systems like a Pro with Mayank Shrivastava
https://t.co/dNAH5d6PB5

Learnings from building Open Source Distributed Systems with Kishore Gopalakrishna
https://t.co/sMgbzTNjFF

[Mayank Shrivastava @mayankshriv]

@Singaporeair_ir Thanks, I have dm’ed the details. @Singaporeair_ir

[Mayank Shrivastava @mayankshriv]

@SingaporeAir Didn't really expect this from you @SingaporeAir. You are making us fly from another city, get stranded for 12 hours, lose what we paid for connecting flights and this is the response.

[Mayank Shrivastava @mayankshriv]

@SingaporeAir Thank you, looking forward to it.

[StarTree @startreedata]

Welcome to Apache Pinot 201! Your video curriculum on the advanced topics of @ApachePinot 📹

Join @gAmUssA, our Head of Developer Advocacy, for this 11-part series that will enhance your skills and empower you with practical, actionable insights: https://t.co/Wv836xD6VV

[Mayank Shrivastava @mayankshriv]

@bztree @lemire @ApachePinot @richardstartin @startreedata Spilling to disk is ok but mostly useless. The expectation of applications using Pinot is sub second/interactive. Spilling to disk is of not much use because the timeout is much smaller.

[Mayank Shrivastava @mayankshriv]

@bztree @lemire @ApachePinot @richardstartin @startreedata Pinot does return a flag in the response indicating that the results are partial. Users see that and increase the limit. We understand where you are coming from and we had the same opinion when we built Pinot initially but users preferred this behavior vs failing the query.

[Mayank Shrivastava @mayankshriv]

@bztree @lemire @ApachePinot @richardstartin @startreedata Also, most practical use cases have an order by clause with group by, and several of them are mostly interested in the TOP N results. It's a popular heuristic in distributed systems to process topN * K from each worker, to improve speed without sacrificing accuracy.

[Mayank Shrivastava @mayankshriv]

@bztree @lemire @ApachePinot @richardstartin @startreedata The behavior to limit groups is configurable. The configurability allows applications to tune the behavior as per their requirements in terms of “accuracy” vs “speed”.  Whereas, the default behavior behavior protects the system against bad queries.

[StarTree @startreedata]

Insights from our CEO, @KishoreBytes, on our last year and what's ahead for StarTree 📰 ⬇️

2023 was an eventful year that paved the way for what's ahead. Read some notes from our Co-Founder and CEO to learn more: https://t.co/wfL0W8DcMu

[StarTree @startreedata]

Wikipedia is changing at a rate of 2 edits per second 🤯

Join @hkdulay to visualize those changes, leveraging Upserts and Ingestion Transformation to get your data into @ApachePinot before visualizing it with Jupyter Notebook 🪐

Register: https://t.co/igVGpSC7pW

[StarTree @startreedata]

💥 Want StarTree + @ApachePinot SWAG and take home an iPad? Find us at AWS re:Invent booth #M6B! 🎁

Join us in the Marketplace Pavilion to learn from our real-time analytics experts and add some gadgets to your swag collection. Learn more: https://t.co/LWU6A4V12R

[StarTree @startreedata]

⭐ Innovators & thought leaders - the StarTree All-Stars Class of 2024 awaits your application, open through November.

Elevate the @ApachePinot & StarTree experience—gain early feature access & exclusive swag! Apply or nominate a community star today: https://t.co/DIsxP0X6xN

[Mayank Shrivastava @mayankshriv]

Impact indeed, I'd say. An excellent blog by @UberEng on the impact created using @ApachePinot. https://t.co/kBWDWl5YAY

[Uber Engineering @UberEng]

Updating enterprise-scale software & infra without creating downstream issues is a daunting task–we leveraged Apache Pinot to create Healthline, a tool that uses real-time & historical crash analytics to minimize release errors in real time & @Uber scale.

https://t.co/EGEmIkxNyS

[StarTree @startreedata]

9/ Over the years, more than three hundred people have contributed to the project and produced over 1.3 million source lines of code. Here are some of the most active contributors: @JackieXiaotian, @xiangfu0, @alexanderpucher, @mayankshriv, @nehapawar18, @richbird9 and many more.

[StarTree @startreedata]

Morning San Francisco, we are excited to return for Day 3 of #DataAISummit!

Booth 47 is the place to talk about real-time analytics, grab some cool @ApachePinot swag, enter our Apple Watch raffle, and sign up for a StarTree Cloud Free Trial. We're excited to see you there💥

[StarTree @startreedata]

Food delivery service @letsblinkit is running an event-driven infrastructure powered by Apache Pinot™.

At a recent StarTree Huddle, they shared how their event collection framework powers observability across the Blinkit platform: https://t.co/8wdpzymELX

[StarTree @startreedata]

In the latest episode of the Real-Time Analytics Podcast, @tlberglund is joined by @mayankshriv to discuss the historical developments of @ApachePinot as it moved along.
Watch the full episode at https://t.co/z9ZsNQLryj or in your favorite podcast player https://t.co/BSIBB8dYMb