Sign in to view Marina’s full profile
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
New York, New York, United States
Contact Info
Sign in to view Marina’s full profile
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
347 followers
348 connections
Sign in to view Marina’s full profile
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
View mutual connections with Marina
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
View mutual connections with Marina
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Sign in to view Marina’s full profile
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Activity
Sign in to view Marina’s full profile
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
-
Playing tennis with Roger Federer was one of my favorite parts of 2017. I can’t wait to team up with him again on March 5th for the next Match for…
Playing tennis with Roger Federer was one of my favorite parts of 2017. I can’t wait to team up with him again on March 5th for the next Match for…
Liked by Marina Ciocea
-
Skype wins an Emmy award! Who can forget the news interview interrupted by the children :D Great job team!!!!!!!!!!!
Skype wins an Emmy award! Who can forget the news interview interrupted by the children :D Great job team!!!!!!!!!!!
Liked by Marina Ciocea
Experience & Education
-
Google
******** ********
-
*********
******** ********
-
****
****** ******** ******** *
-
***** ************ *********
****** ** ******* (*.**.) *** **** ******** *** *********** ******** *******
-
-
************* „***********” *** ******ș**
****** ** ******* (*.**.) *** **** ******** *** *********** ******** *******
-
View Marina’s full experience
See their title, tenure and more.
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View Marina’s full profile
Sign in
Stay updated on your professional world
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Other similar profiles
-
Leanne Ogasawara
--
JapanConnect -
Emma (Mirică) Goșu
Greater Stockholm Metropolitan AreaConnect -
Dalibor Kukoleca
Greater Stockholm Metropolitan AreaConnect -
Jossif Tomash
Release Manager
StockholmConnect -
Emilian Miron
United StatesConnect -
Valentin Gosu
Software Engineer at Mozilla
StockholmConnect -
Claudiu Coman
Bucharest Metropolitan AreaConnect -
Linus Upson
San Francisco Bay AreaConnect -
Allan Cass
Electrical Engineer at NA
Drayden, MDConnect -
Glen Davis
at
Richmond, VAConnect -
Vanessa Diaconu
--
Toronto, ONConnect -
Vanessa Diaconu
Subject Matter Expert la Genpact
Popeşti-LeordeniConnect -
Vanessa Diaconu
--
RomaniaConnect -
vanesa diaconu
Manager la Creative Artistik
RomaniaConnect -
Jakob Engstrand
Books. Music. Photography.
UppsalaConnect -
Christopher Baisan
LondonConnect -
Mihaela Dvornic
RomeConnect -
Luan Hewson
Empresário RED Wolf Brazil
Foz do Iguaçu, PRConnect -
Daniel Johnson
Software Engineer at Google Inc.
San Francisco Bay AreaConnect -
Emily Soldal Norelius
StockholmConnect
Explore more posts
-
Colt McNealy
Just a reminder that I'll be talking about how to safely share a Kafka cluster among many users in an organization using #strimzi at the virtual Strimzi Conference on May 22! I'm very excited about this one. You'll learn: - Kafka Topic Management with the `KafkaTopic` CRD - Creating Kafka Principals using the `KafkaUser` CRD - How Quotas work in Apache Kafka - How and why to rebalance clusters using Strimzi's Cruise Control integration - How all of this applies for the upcoming LittleHorse Cloud Enterprise service #apachekafka #conference Sign up here: https://lnkd.in/gm_Pbhsg
21 -
Bo Ingram
When I was first thinking about what I wanted ScyllaDB in Action to be, I knew I didn't want to stop at the usual features; I wanted a book that guided you through running the database in production. In the newly-available chapters 8 and 9 of ScyllaDB in Action, you'll learn about the internal systems that make ScyllaDB work, giving you a solid foundation to reason about the database's behavior as you then study how to configure and work with a production ScyllaDB cluster. http://mng.bz/46Jw ScyllaDB Manning Early Access Program (MEAP)
762 Comments -
Valerio Gheri
Have you ever wondered whether the order of fields within a Go struct affects memory consumption or application performance? I did, and it turns out that it might indeed have an impact. The reason? Memory alignment requirements in modern CPU architectures. I documented my understanding of the topic at https://lnkd.in/ddDsG8MU . Take a look if you're interested, feedback is appreciated!
13 -
Colt McNealy
<rant subject = #streaming and #queues> Streams have stronger ordering guarantees than message queues. Generally this means that all messages within a partition are read in the same order. This is highly useful when you need an ordered record of events that happened. As a consequence, streaming systems don't have individual message acknowledgement as it would violate ordering guarantees. Message queues have per-message acknowledgement, which means they have less strict ordering guarantees. One hybrid approach which is quite cool is "key-based subscriptions" in Apache Pulsar and also Confluent's Parallel Consumer for #apachekafka which guarantee order for all messages with the same key while allowing you to individually acknowledge messages. The tradeoff with individual message ack's is persistence (disk or memory)—you need to keep track of which messages have been acknowledged and which ones are still inflight. Pulsar uses BookKeeper to persist "subscriptions." Parallel Consumer stores them in the user-defined metadata in the consumer offset topic. There's an in-flight KIP for "Queues on Kafka" (KIP-932) but if I were the Grumpy Maintainer of Kafka, I would say that it's unnecessary given that it's already possible to implement queue semantics on top of Kafka in the client side (see: Parallel Consumer). Why add additional code bloat on the broker side when the features are already accessible? </rant>
4 -
Jeremy Manson
In one of my recent posts complaining about software engineering practices, I mentioned that a good strategy for encouraging quality is to make sure that developers are steered in the right direction by their development environment - the phrase I used was (approximately) to "make the easy thing also be the right thing". Some folks chimed in and said that the right thing often costs too much upfront. That was precisely my point - if it costs too much upfront, it's too hard. The right thing should be easy, not hard! It turns out that - O cruel fate! - it takes a lot of hard work to make the right thing also be the easy thing. One of my favorite papers on this topic came from folks in my orbit at Google a few years ago. Members of my team had done a lot of work on static analysis for Java, and worked closely with our colleagues in engineering productivity to write up the lessons learned about what makes a good static analysis tool. The entire article is worth reading, but spoilers: after a lot of case studies, they identified that static analysis tools must be: Integrated. Developers will not go out of their way to run these tools, so the tool must be part of the developer's workflow, and not take too long to run. Have actionable errors. Developers ignore incomprehensible errors, as well as errors for which the followup action isn't clear. Ideally, the tool will tell you exactly what to do. Trustworthy. A tool cannot have too many false positives. For a static analysis tool to be included in our default code review workflow at Google, developers should feel the check is pointing out an actual issue at least 90% of the time. The error must manifest in practice. If a developer knows that a bug will never actually happen, they are unlikely to fix it. The fix must be cost-effective. Bugs that require you to rewrite your code from scratch must be very important. All of this requires a strong partnership between the static analysis tool developer and the user. One of the things I have enjoyed most about working at Google is that that kind of partnership is possible - Google has one big multi-billion line codebase, with one set of compilers and static analysis tools that we all share. The fact that Google has embraced this philosophy for static analysis tools has meant that they are widely embraced, adopted, and seen as a win. Side-note: One of the things that I've most liked about moving to Rust is that several of these lessons are baked into the compiler. It is very, very common to find that the Rust compiler knows exactly what change you need to make to your code, and tells you. For example, I find it very easy to forget whether I'm dealing with a reference or a value in Rust, and the compiler will literally spit out a diff for me. Sometimes the error messages could be improved, but, by and large, I've been pleased to see their embrace of this approach. The entire article is worth reading. https://lnkd.in/dJhCXBSE
21 -
Vlad Leyberov
Google SRE NYC Tech Talk - May 22nd, 6:30pm Google SRE NYC proudly announces our second Tech Talk event in 2024. The in-person only event will take place on Wednesday, 22nd of May 2024 at 6 PM at our Chelsea Markets office in NYC. The doors will open at 5:30 pm. Spaces are limited, RSVP via our Meetup event page. https://lnkd.in/ef5fqdrE Agenda: Salvatore Furino - Customer Reliability Engineer (CRE) at Bloomberg “The Hammer Changes the Hand” This talk will briefly explore how to view internal tooling through the lens of product management in not just developing and shipping features, but how those features empower teams to change their understanding of their social-technical systems. Sal is a Customer Reliability Engineer. During his career he’s worked as a TPM, SRE, Developer, Sys Admin, and IT support. While not working he enjoys cooking, gaming, traveling, skiing, and golfing. Sal lives in Queens and has a BS in Applied Mathematics from Marist College. Thiara Ortiz - Staff CDN Reliability Engineer at Netflix “How we measure Quality of Experience to ensure our members get a world class experience they have come to expect from Netflix ” Any time a Netflix member sits down, reclines in their chair and turns on their TV to Netflix, there's a moment of truth. It's an opportunity to deliver a spectacular service with amazing quality of experience. This talk will go over how we measure the quality of experience for our members and how we work to develop new metrics when we have additional offerings like live streaming and cloud gaming. Thiara has worked at some of the largest internet companies in the world, Meta and Netflix. During her time at Meta. Since Meta, Thiara has been working at Netflix as a Staff CDN Reliability engineer. Her focus is primarily on resilience and quality of experience for members streaming from Open Connect. When incidents occur and Netflix's systems do not behave as expected, Thiara can be found working and engaging the necessary teams to remediate these issues. Mike Scherbakov - Staff Site Reliability Engineer at Google “LLM for SRE / Using LLM in SRE space” LLMs open up an opportunity to automate and scale many operational processes, which couldn't be otherwise solved by conventional methods. Examples include simple summarization of issues and incidents, assisting production on-callers, managing incidents, clustering (creating taxonomy) of issues, scaling SRE via assisted review of development design documents. Therefore LLMs provide a new and unique opportunity to transform the work we do as SREs. Mike works in YouTube Ads SRE as well as co-leading LLM4SRE, LLM center of competency in SRE at Google Our Tech Talks series are focused on professional development and networking: no recruiters, sales or press are allowed.
163 Comments -
Todd Nist
pgvectorscale: A Game-Changer for Vector Similarity Search 🚀 Timescale's #pgvectorscale extension for PostgreSQL has achieved a remarkable milestone, matching the performance of Pinecone, a dedicated vector database, while being 75% more cost-effective. 💰 In their latest blog post, Timescale showcases the impressive speed and efficiency of pgvectorscale, which leverages the power of PostgreSQL to deliver lightning-fast vector similarity search. This breakthrough opens up new possibilities for applications like semantic search, recommendation systems, and more. 🔍 Key takeaways: 🌟 pgvectorscale now matches Pinecone's performance in vector similarity search 💸 pgvectorscale is 75% more cost-effective compared to Pinecone 🔧 Easy integration with existing PostgreSQL databases 🌐 Scalable solution for handling billions of vectors ⚡ Optimized for fast and efficient search results pgvectorscale's seamless integration with PostgreSQL allows developers to incorporate vector similarity search into their applications with ease, while benefiting from the robustness and scalability of a proven database system. 🗄️ If you're looking for a high-performance, cost-effective solution for vector similarity search, pgvectorscale is definitely worth exploring. 🔍💡 #pgvectorscale #postgresql #vectorsimilaritysearch #semanticsearch #recommendationsystems #timescale https://lnkd.in/gsfgCK32
5 -
Ian Cooper
To lift my reply up, I have expanded it here. These topics form much of the material on Day 2 of my Practical Messaging workshop next NDC Oslo 10-11 June) So the first thing to think about is the type of asynchronous conversation: * Messaging (uses a queue; transfer of control) * Eventing (uses a stream; publish-subscribe) (Replies can be either stream or queue, default to queue unless you are broadcasting). We also need to think about Provider over Producer/Consumer when thinking about the interaction. Who provides the API? The Consumer provides the API for messaging; the Producer provides the API for eventing. Concepts like In-Out, Out-Only are useful here from the old WS- Messaging specs. I use that taxonomy when I teach my workshop because it is still a useful tool. For example in In-Out we have a consumer that acts as the API provider and receives a message; it should use a queue; we have a response that probably travels back over that queue. The main issue that brings this up is causality - I need this step (taking a payment) to occur before this step (place an order). Definitely seen a few systems struggle with “event chaining” where they lose causality, particularly pernicious when a component fails (and can lead to nasty loops where work goes round and around - imagine folks keep turning up with the same order to your house). I spend time in my trying to avoid everyone reaching for a Process Manager (Saga) here. In a lot of cases I recommend what I call Turn-Taking which is essentially Pat Helland’s Activity model (components co-operate via their internal business logic).
314 Comments -
Phil Strong
Blueprints are templates you can apply to one or more code repos to introduce best practices or jumpstart a new project. Later, you can update the blueprint and apply those changes, like a new version of Node or Log4J, to all repositories using the template. This CodeCatalyst feature is now extended to GitHub repositories!
7 -
Leonid Gorbachev
Last Saturday, my 8 years-old son with help of Amazon CodeWhisperer created his primitive videogame about astronaut that collects rocks on the moon. The idea: The Moon, moonrocks are scattered randomly on the moon, the astronaut on the moon is moved by WASD keys, SPACE - collect a rock, collected rock appears in free cell of inventory, T - opens a table where you can convert four rocks into a house. We used: VSCode with AmazonQ extension, python3.12. Step by step, he used around 6-7 prompts in AmazonQ chat to create what he wants, without any knowledge of python. At one moment, i had to step-in because AmazonQ couldn't fix himself after 3 fix-prompts, but it was because the kid introduced a complexity that nobody expected - he asked for 5 different types of rocks that must be handled properly in the game. Of course, few days before i checked by myself - will it be possible to get some valuable result or not to do not upset the kid. And I would say that you can get applicable results in ~5 prompts without manual intervention or with ~3 prompts with small human changes or fixes. Also i compared with Microsoft Copilot (free license) with using of absolutely the same prompts - at the very beginning step, MSCopilot gave the same result, but after 3rd prompt it lost the context of dialogue and unexpectedly started to generate a pictures instead of python code. It will be interesting to try the same with Github Copilot. The result: a kids from primary classes can create a simple games without any knowledge about syntax, types, tuples and etc, only with using of AI. And then he can start to learn on working example - what is events, what is resolution and objects and how it all works.
392 Comments -
Peter Warnock
Solid article from Charity Majors of honeycomb.io. A few key takeaways resonated with me: - Automation and GenAI are tools, not replacements for human expertise. We need to understand how to use them to augment our work, not do it for us. We can do more, better and faster, when we learn to leverage the utility of the tools. - The lack of mentorship in our industry is a disservice to junior engineers. We should be building our teams like a farm system in baseball, investing time and energy in growing and nurturing talent. It builds a sustainable organization. - The 7-year mark for achieving senior-level mastery is a realistic goal. It's a reminder that expertise takes time and dedication, echoing the 10,000 hour rule popularized by Malcolm Gladwell. -Charity's article is a timely reminder that there's no shortcut to building a strong engineering team. It takes effort, investment, and a commitment to growing and developing our people. What are your thoughts on the role of automation and mentorship in engineering teams? Share your insights in the comments below!
6 -
Ahsan Nabi Dar
Large Language Models (#LLMs) are evolving rapidly, and the demand for running them at scale is skyrocketing. #Ollama is stepping up to meet this need with its official container image on Docker Hub, excelling in local #deployments. However, there’s a catch for containerized environments: downloading #models requires the ollama serve command to be running within the #container. This extra step complicates automated deployments via #CI #CD #pipelines, making seamless model access challenging. To take your #model to production, it needs to be #container-ready—a feature Ollama has yet to fully support. The gap in #AIOps for smooth deployments gets blurred with the inference services available as an easy way to take it to production. Goal is to overcome obstacles and deliver solutions 😉 so here is to get you started on having your (O)#llama #production ready in a private #deployment from your #local to #cloud all in #oneshot . Enjoy Medium: https://lnkd.in/g9jqBQ8d dev.to: https://lnkd.in/gi_gGwam #AI #LLM #Ollama #llama3 #phi3 #mistral #S6overlay #AIOps #CI #CD #oneshot #container #docker #podmanEnjoy
11 -
Yacine Taleb
Our SIGMOD '24 paper explores MemoryDB's architecture, which delivers in-memory performance and strong consistency at scale. A key innovation is the decoupling of durability and replication from the in-memory database engine, which allows scaling compute and storage independently based on customer workload needs. Customer data integrity, security, and availability are paramount for MemoryDB. The paper dives deep into innovations that the team built to make that possible (a personal favorite is off-box snapshotting which allows us to create consistent database snapshots on separate clusters to avoid impacting live customer traffic and preventing resource contention). We also share hardened operational experiences from running a distributed database at cloud scale. #aws #memorydb #inmemory #databases #durability #sigmod
911 Comment -
Logan Abbott
The Legendary Zilog Z80 CPU Is Being Discontinued After Nearly 50 Years: Long-time Slashdot reader AmiMoJo shares an article from TechSpot: Zilog is retiring the Z80 after 48 years on the market. Originally developed as a project stemming from the Intel 8080, it eventually rose to become one of the most popular and widely used 8-bit CPUs in both gaming and general computing devices. The iconic IC device, developed by Federico Faggin, will soon be phased out, and interested parties only have a few months left to place their orders before Zilog's manufacturing partner ends support for the technology... Federico Faggin, an Intel engineer, founded Zilog in 1974 after his work on the Intel 4004, the first 4-bit CPU. The Zilog Z80 was then released in July 1976, conceived as a software-compatible 'extension' and enhancement of the Intel 8080 processor. Back in 1999 Slashdot was calling Zilog's updated eZ80 "one of the fastest 8-bit CPUs available today, executing code 4 times faster than a standard Z80 operating at the same clock speed." Another headline, from 2001: Zilog To File For Chapter 11... Read more of this story at Slashdot.
-
Jignesh (Jiggy) Kakkad
Wow, integrating two different LLMs and making them work together seamlessly is truly impressive! It's no easy feat ensuring that both prompts and parsing work in harmony, but when it's done successfully, it results in a system that's more efficient and effective. You can count on accurate and reliable responses that'll make your day-to-day tasks much easier. #PromptEngineering #RAG #LLMs #Gemini Vineet Kumar Maheshwari
161 Comment -
Ashish Dhar
Exciting news for AWS CloudFormation users! Introducing the new ServiceTimeout property for custom resources, enabling you to set a maximum timeout for provisioning logic execution. This feature accelerates the dev-test cycle, providing quicker feedback. Previously, custom resource timeouts were fixed at 1 hour, causing delays in error situations. Now, with customizable timeouts, execution efficiency is enhanced, improving development and testing iterations. #aws #cloudformation https://lnkd.in/gTaSmMhM
141 Comment
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More