Archives
-
Ready to Expand into the U.S.? Learn from the Experts!
Ready to Expand into the U.S.? Learn from the Experts!
Hi MD,
Ready to take your business global?
Expanding into the U.S. is a pivotal move for international companies seeking growth, but it comes with a unique set of challenges. Join us to discover how leading firms are setting up for success in one of the world’s most dynamic markets.
Webinar: Money, money, money! Raise, bank, pay, and grow in the USA
Join J.P. Morgan, Wilson Sonsini, and Remote as they reveal the critical steps for international expansion into the U.S.
🗓 Date: Thursday, December 5, 2024
⏰ Time: 4pm UTC
What you’ll gain:
- Raise Capital – Discover strategic fundraising solutions to support your U.S. growth.
- Payroll Essentials – Set up your workforce smoothly and ensure compliance from day one.
- Legal Compliance 101 – Navigate U.S. regulations with ease to avoid costly missteps.
- Banking Strategies for Growth – Explore funding solutions tailored for international expansion.
Don’t miss out on this opportunity to unlock insights from top industry experts and make your U.S. expansion a success!
Speakers
Barry Flanagan
Vice President, Payroll at Remote
With over a decade of experience as a global payroll executive, Barry is an expert in helping businesses build scalable payroll solutions and ensuring customer satisfaction.
Daniel Glazer
London Office Managing partner at Wilson Sonsini
Daniel heads the US Expansion group at Wilson Sonsini, specializing in guiding UK and EU tech startups through the US lifecycle, from fundraising to M&A/IPO.
Barry O’Brien
Head of Global Passport, Innovation Economy at J.P. Morgan
Barry leads J.P. Morgan’s Global Passport team, supporting venture-backed startups and growth companies with tailored banking solutions for US expansion.
What are you waiting for?
This is your chance to get expert guidance on navigating US expansion successfully. Join us to set your team up for success!
You received this email because you are subscribed to News & Offers from Remote Europe Holding B.V
Update your email preferences to choose the types of emails you receive.
Unsubscribe from all future emailsRemote Europe Holding B.V
Copyright © 2024 All rights reserved.
Kraijenhoffstraat 137A 1018RG Amsterdam The Netherlands
by "Remote" <hello@remote-comms.com> - 04:00 - 20 Nov 2024 -
The energy transition: Where are we, really?
Only McKinsey Perspectives
by "Only McKinsey Perspectives" <publishing@email.mckinsey.com> - 01:42 - 20 Nov 2024 -
GLA Project Network - The best choice to expand business
Dear Partner,
Good day.
Greeting from Ms. Joy and GLA family.
Whether your company focus on project logistics services, such as Bulk Cargo, RO-RO, Break Bulk, Heavy Lift etc. ?
Any further details, keep us posted.
Thank you,Conference register: http://www.glafamily.com/meeting/registration/Joy
(Notice: If you would not like to receive email blush, please inform us timely. We will delete your data from the database)
Joy Jiang
Overseas Dept
M: +86 153 6164 8088;
( Whatsapp & Wechat)
A:GLA Co.,Limited
HongChang Plaza, 2001,Shennan Dong Road,Shenzhen
518000,China
by "GLA family" <member321@glafamily.com> - 10:24 - 19 Nov 2024 -
Don't Miss Out: Dreamforce to You webinar happening next week
November
27 11:00 SGT
Live webinar
MuleSoft: Dreamforce 2024 to You
Register now Join us on 27 November 2024, 11am SGT for an exclusive Dreamforce to You webinar focused on MuleSoft. Discover our latest product innovations and equip your teams with advanced tools and knowledge.
Learn how to:
- Build an AI-powered composable foundation
- Explore Agentforce's capabilities
- Integrate and automate seamlessly
- Develop orchestrated workflows
Speakers
Andrew Comstock
VP of Product Management
Salesforce
Jing Li
Senior Director of Product Management
Salesforce
Mofeyi Oluwalana
Senior Product Manager
Salesforce© 2024, Salesforce, Inc.
Salesforce.com 2 Silom Edge, 14th Floor, Unit S14001-S14007, Silom Road, Suriyawong, Bangrak, Bangkok 10500
General Enquiries: +66 2 430 4323
This email was sent to info@learn.odoo.com
Manage Preferences or Unsubscribe | Privacy Statement
Powered by Salesforce Marketing Cloud
by "Salesforce Webinars" <salesforce@mail.salesforce.com> - 09:32 - 19 Nov 2024 -
Patches Brodés | Patches Chenille | Patches en Cuir et Bien Plus Encore
Bonne Journée,Je m'appelle Donna. Nous avons réuni une équipe de spécialistes qualifiés, dynamiques et engagés pour créer les meilleurs projets possibles pour nos clients. Nous proposons désormais une gamme de prix compétitive sur le marché.
Nous ne facturons pas de devis, alors n’hésitez pas à nous envoyer vos créations pour obtenir un devis.
Types de Patches que Nous Réalisons :
- Patches Brodés
- Patches en PVC Caoutchouc
- Porte-clés Brodés
- Patches de Nom
- Patches Woburn avec Sublimation
- Patches Tissés
- Patches Appliqués
- Patches Imprimés
- Patches en Cuir
- Patches Chenille
- Étiquettes
- Dos Thermocollant
- Dos Velcro
- Dos à Coudre
POURQUOI USA DIGITIZING LLC ?
- Patches Brodés Personnalisés de Haute Qualité
- Délai de Production le Plus Rapide : 5-7 Jours Ouvrables
- Formats de Fichiers Illimités
- Prix Raisonnables
- Meilleure Qualité
- Révisions Illimitées Gratuites
- Support Client Premium 24/7
Nous vous assurons que vous recevrez des produits de haute qualité et le meilleur patch possible en fonction de vos besoins.
Ensemble, Nous Pouvons Grandir !
Donna Harper Sales Executive www.usadigit.com custompatches@usadigit.com +1 717 962 9443 802 Evergreen Ct North Wales, PA, 19454,US 
by "USA Embroidery | Chenille | Leather Patches" <custompatches@usadigit.com> - 08:09 - 19 Nov 2024 -
How to deliver effective performance reviews
Intersection
Share this email
Companies that prioritize their people’s performance set up their employees—and their organizations—for success, say McKinsey partner Asmus Komm and coauthors. To learn more about how companies can get the most out of their year-end review process, check out the latest edition of the Five Fifty.
We’re commemorating our flagship publication’s milestone birthday with a year of themed issues, special interactive features, and timeless Quarterly classics newly released from the archives. It’s all available with a free digital Quarterly membership. Sign up here.
Share these insights
Did you enjoy this newsletter? Forward it to colleagues and friends so they can subscribe too. Was this issue forwarded to you? Sign up for it and sample our 40+ other free email subscriptions here.
This email contains information about McKinsey’s research, insights, services, or events. By opening our emails or clicking on links, you agree to our use of cookies and web tracking technology. For more information on how we use and protect your information, please review our privacy policy.
You received this email because you subscribed to our McKinsey Quarterly Five Fifty alert list.
Copyright © 2024 | McKinsey & Company, 3 World Trade Center, 175 Greenwich Street, New York, NY 10007
by "McKinsey Quarterly Five Fifty" <publishing@email.mckinsey.com> - 05:27 - 19 Nov 2024 -
Uber’s Billion Trips Migration Setup with Zero Downtime
Uber’s Billion Trips Migration Setup with Zero Downtime
Forwarded this email? Subscribe here for more📅Meet your EOY deadlines – faster releases, zero quality compromises (Sponsored)
If slow QA processes and flaky tests are a bottleneck for your engineering team, you need QA Wolf.
Their AI-native platform, backed by full-time QA engineers, enables their team to create tests 5x faster than anyone else. New tests are created in minutes and existing tests are updated almost instantaneously.
✔️Unlimited parallel test runs
✔️15-min QA cycles
✔️24-hour maintenance and on-demand test creation
✔️CI/CD integration
✔️Zero-flake guarantee🛒Also available through AWS, GCP, and Azure marketplaces.
Disclaimer: The details in this post have been derived from the Uber Technical Blog. All credit for the technical details goes to the Uber engineering team. The links to the original articles and other references are present in the references section at the end of the post. We’ve attempted to analyze the details and provide our input about them. If you find any inaccuracies or omissions, please leave a comment, and we will do our best to fix them.
Maintaining uptime during system upgrades or migrations is essential, especially for high-stakes, real-time platforms like Uber's trip fulfillment system.
Uber relies on consistent, immediate responsiveness to manage millions of trip requests and transactions daily. The implications of downtime for a platform like Uber are scary: even a brief service interruption could lead to lost revenue, user dissatisfaction, and huge reputational damage.
However, Uber faced the daunting challenge of migrating its complex fulfillment infrastructure from an on-premises setup to a hybrid cloud environment.
This task not only required a deep understanding of system architecture but also an approach that could guarantee zero downtime, ensuring that millions of riders, drivers, and partners worldwide would experience uninterrupted service during the transition.
In this article, we’ll look at Uber's zero-downtime migration strategy. We’ll also learn more about the technical solutions they implemented and the challenges they faced in the process.
The Complexity of Uber’s Trip Fulfillment Platform
Uber’s fulfillment system, before it migrated to a hybrid cloud setup, was designed to handle vast amounts of real-time data and high transaction volumes.
At its core, the fulfillment system managed the interactions between millions of riders, drivers, couriers, and other service elements, processing over two million transactions per second. Some stats to describe the system’s importance are as follows:
The platform handles more than a million concurrent users and billions of trips per year across over ten thousand cities.
The platform supports billions of database transactions a day.
Hundreds of Uber microservices rely on the platform as the source of truth for the accurate state of the trips.
The architecture was organized into “pods”, where each pod was a self-contained unit of services dedicated to a particular city or geographic region. See the diagram below:
For reference, a pod is a self-sufficient unit with many services interacting with each other to handle fulfillment for a single city. Once a request enters a pod, it remains within the services in the pod unless it requires access to data from services outside the pod.
Within each pod, services were divided into “demand” and “supply” systems. The demand and supply services were shared-nothing microservices with the entities stored in Apache Cassandra and Redis key-value tables.
The demand services managed rider interactions
Supply services handled driver operations.
These services stayed synchronized using distributed transactional techniques to ensure that trip-related data remained consistent within each pod. Uber employed the saga pattern for this.
Entity consistency within each service was managed through in-memory data management and serialization.
Architectural Limitations and Scalability Issues
The initial architecture was designed to prioritize availability, sacrificing some aspects of strict data consistency to maintain a robust user experience.
However, as Uber’s operations expanded, these architectural choices created bottlenecks:
Eventual Consistency: The entire architecture was based on the thought process of trading off consistency for availability and latency. The lack of atomicity meant the need to reconcile when the second operation failed.
Multi-Entity Write Operations: When an operation had to write across multiple entities, the application layer handled this interaction using an RPC-based mechanism. The system had to constantly verify the expected and current states to fix mismatches.
Limited Scalability: The cities were sharded among one of the available pods and the size of the pod was dependent on the maximum ring size of a cluster. There was a vertical limit for scaling the pod if any of the cities crossed a threshold of concurrent trips.
Learn the Roadmap to making $100k using LinkedIn & AI 🚀(Sponsored)
LinkedIn isn’t just a social platform—it’s a goldmine when you combine it with AI.
In his AI Powered LinkedIn Workshop, you will learn how to harness the power of LinkedIn as a founder, marketer, business owner, or salaried professional.
In this workshop, you will learn about how to:
👉 Automate lead generation to grow your business while you sleep
👉 Leverage AI to land high-paying jobs without wasting hours on application
👉 Master his $100K LinkedIn Outbound Strategy to boost revenue effortlessly
👉 Use AI to create and distribute content, saving you hours every week
This workshop is the real deal for anyone who wants to dominate LinkedIn in 2024 and beyond.But it’s only FREE for the first 100 people. After that, the price jumps back to $399.
The Redesigned Architecture
In the redesigned system architecture, the Uber engineering team shifted from a distributed, in-memory setup to a more centralized, cloud-backed infrastructure.
The new system consolidated the previously separate “demand” and “supply” services into a single application layer supported by a cloud database. By moving data management to a datastore layer, the new system streamlined operations and improved scalability and consistency.
See the diagram below to understand the new architecture:
Key Solutions Implemented By Uber
While designing the new solution had its share of the complexity, the real challenge was migrating the workload to the new design.
Some of the key solutions implemented by Uber’s engineering team to achieve zero downtime migration to the new system are as follows:
1 - Backward Compatibility Layer
Uber implemented a backward compatibility layer as a core component of its zero-downtime migration strategy.
This layer served as a bridge, allowing existing APIs and event contracts to function normally while Uber transitioned to a new system architecture. By supporting the old API contracts and event schemas, the backward compatibility layer ensured that many internal and external consumers of Uber’s APIs could continue to operate without modification.
See the diagram below to understand the role of the backward compatibility layer.
Some of the benefits of the backward compatibility layer are as follows:
By keeping old API contracts intact, Uber avoided abrupt changes that could cause errors or service interruptions, ensuring a stable user experience throughout the migration.
Consumers could move to the new system at their own pace, enabling a gradual adoption of the new API endpoints without disrupting workflows.
The compatibility layer minimized the need for coordination across all API consumers, as teams could adopt the new system independently.
However, there were also some downsides to this:
Increased Complexity: Maintaining a compatibility layer added another source of complexity to the migration. Every API endpoint and event schema had to be kept consistent across both systems, leading to possible redundancy in data processing and storage.
Risk of Technical Debt: Continuing to support outdated APIs and event schemas can create long-term technical debt. If not carefully managed, the compatibility layer could lead to ongoing maintenance costs and slow down future development.
Performance Overheads: The compatibility layer could introduce performance overhead, as requests may need to be processed in both the old and new systems or converted to fit the new architecture. This can result in latency, especially in high-volume systems.
2 - Shadow Validation
Shadow validation was integrated into Uber’s high-transaction, real-time platform.
Each request sent to the old system was mirrored in the new one, and responses from both were compared on a key-value basis. Discrepancies were logged and analyzed within Uber’s observability framework, with differences captured in a dedicated observability system for further examination.
This comparison was not simply binary; Uber allowed for certain predefined tolerances and exceptions to accommodate unavoidable variances, such as transient data changes or slight delays due to processing orders.
When differences between systems arose, Uber took a two-fold approach:
Automated Logging and Alerts: Minor discrepancies within pre-approved tolerances were logged for further analysis but did not require immediate action. These were typically edge cases where slight timing differences or transient state mismatches occurred.
Manual Intervention for Critical Mismatches: If a response deviation posed a risk to user functionality or violated critical data requirements, the engineering team could prioritize it for immediate investigation. This step ensured that high-impact discrepancies were resolved quickly, reducing the chance of significant issues when the system went live.
Migration Phase
Let’s now understand how Uber carried out the migration.
1 - Pre-Rollout Preparations
Uber’s pre-rollout preparation steps were foundational to achieving a smooth, zero-downtime migration for their trip fulfillment platform.
Each step, such as shadow validation, end-to-end (E2E) testing, load testing, and database warm-up, played a critical role in minimizing potential risks and ensuring the new system would perform reliably under real-world conditions.
Some of the key activities that were performed are as follows:
End-to-End (E2E) Testing
E2E testing enabled Uber to verify the complete functionality of the new system from start to finish. It was essential to identify potential integration issues or bottlenecks.
By simulating realistic user journeys, these tests checked that all workflows, integrations, and dependencies performed as expected under different scenarios.
Load Testing
Load testing subjected the new system to simulated traffic levels that matched or exceeded actual usage, evaluating its capacity to handle high transaction volumes without degradation.
Load testing confirmed that the new system could withstand Uber’s high operational demands, from peak traffic to unexpected surges. It allowed Uber to preemptively address any performance issues, such as latency or system overload, under stress conditions.
Database Warm-Up
Database warm-up involved generating synthetic data loads to pre-fill caches and split partitions, ensuring the database was primed for full production traffic from the start.
For Uber, whose cloud database had to handle rapid scaling, database warm-up prevented “cold-start” issues by ensuring that common queries and data partitions were already optimized for performance. This step reduced the chance of initial slowdowns or resource bottlenecks during migration.
2 - Traffic Pinning and Phased Rollouts
Uber employed a traffic pinning and phased rollout strategy to migrate specific trips incrementally to the new system to reduce the risk of inconsistencies
This approach allowed Uber to gradually shift parts of its trip fulfillment platform to the new architecture.
Technical Process of Traffic Pinning and Phased Rollout
Traffic pinning ensured that each trip’s data was processed by a single system—either the old or the new—throughout its lifecycle.
This was critical for preventing data fragmentation and ensuring consistent trip updates, as each trip involves multiple interactions, such as driver updates, route changes, and fare calculations.
To achieve this, Uber developed a routing logic to “pin” ongoing trips to the system where they were initiated. Before migration, consumer identifiers for riders and drivers were recorded, enabling Uber to route each interaction related to a given trip back to its origin system, preventing mid-trip transitions that could lead to data mismatches.
This tracking persisted until the trip was completed, after which riders and drivers were gradually transitioned to the new system for future trips.
The diagram below demonstrates the concept of traffic pinning.
Source: Uber’s zero-downtime migration Initially, Uber migrated less critical or idle riders and drivers to the new system, followed by active trips in specific cities. Over time, this phased approach allowed Uber to monitor and control the migration, expanding it to cover larger segments and eventually all active users.
The key benefits of traffic pinning are as follows:
Data Consistency: Traffic pinning ensured that all interactions for a specific trip remained in the same system, preventing data inconsistencies. This consistency is essential for real-time platforms where split-second updates are crucial to the user experience.
Reduced Risk of Errors: By phasing out migrations, Uber could control and monitor each stage, identifying potential issues in a manageable scope. If an error surfaced, Uber could isolate it to a specific set of users or regions, minimizing overall impact.
Increased Stability for Users: Due to this approach, trips in progress weren’t affected by system changes, reducing the likelihood of transaction failures or incomplete updates that could disrupt users.
Key Observability and Rollback Mechanisms
Uber developed detailed dashboards and monitoring tools to track metrics like trip volume, trip completion rates, driver availability, and overall system load.
These dashboards provided visibility into performance and data consistency across both systems, allowing engineers to observe how traffic gradually drained from the old system while increasing in the new one.
The goal was that the overall aggregate metrics should remain flat. Key metrics, including transaction success rates, latency, and error rates, were monitored at city-level granularity to catch any localized disruptions.
These observability tools also enabled Uber to spot irregularities in real-time.
For example, if completion rates dropped in a specific city, engineers could quickly investigate and determine whether the issue originated from the new system. The tools flagged critical problems that could trigger a rollback for specific regions, allowing Uber to maintain stability while continuing the migration elsewhere.
The rollback mechanism was equally important.
Uber could reverse traffic flow for any region back to the old system if metrics indicated significant deviations from expected performance.
Conclusion
Uber’s zero-downtime migration approach highlights key technical strategies for complex, large-scale migrations.
By implementing a backward compatibility layer, Uber maintained service continuity, allowing gradual, flexible transitions for its consumers.
The shadow validation and traffic pinning techniques ensured data consistency and stability, while observability and rollback mechanisms provided real-time insights and controlled reversibility in case of issues. These strategies allowed Uber to minimize user impact while migrating critical components.
Despite its success, Uber’s approach faced challenges, such as the high infrastructure demand of maintaining dual systems and the complexity of managing large-scale observability.
References:
Uber’s Fulfillment Platform: Ground-up Rearchitecture to Accelerate Uber’s Go/Get Strategy
Building Uber’s Fulfillment Platform for Planet-Scale Using Spanner
Uber's Blueprint for Zero-Downtime Migration of Complex Trip Fulfillment Platform
SPONSOR US
Get your product in front of more than 1,000,000 tech professionals.
Our newsletter puts your products and services directly in front of an audience that matters - hundreds of thousands of engineering leaders and senior engineers - who have influence over significant tech decisions and big purchases.
Space Fills Up Fast - Reserve Today
Ad spots typically sell out about 4 weeks in advance. To ensure your ad reaches this influential audience, reserve your space now by emailing sponsorship@bytebytego.com
Like
Comment
Restack
© 2024 ByteByteGo
548 Market Street PMB 72296, San Francisco, CA 94104
Unsubscribe
by "ByteByteGo" <bytebytego@substack.com> - 11:36 - 19 Nov 2024 -
Hi there, is Your Telecoms Consultant looking for a new CRM system?
Hi there, is Your Telecoms Consultant looking for a new CRM system?
Hi there,
With over a decade of experience working with businesses like Your Telecoms Consultant, we focus on designing Customer Relationship Management (CRM) systems, tailored to the unique needs of each company.
Whether you're new to CRM systems or contemplating an upgrade due to system limitations, I urge you to reconsider an off-the-shelf route!
While off-the-shelf CRM system options offer a budget-friendly entry, the costs escalate as your business grows. Opting for a custom system upfront involves a one-time cost, ensuring a tailored fit without ongoing expenses.
Here's why our custom CRM systems stand out:
- Tailored to your needs: Unlike off-the-shelf options, a custom CRM system is built precisely for your business, aligning every feature and functionality perfectly.
- Scalability and flexibility: With a custom CRM system, you have the liberty to scale and adapt as your business evolves.
- Enhanced user experience: Custom CRM systems prioritise user-friendly interfaces and tailored functionalities, easing the onboarding process for your team.
- Greater control and ownership: With our custom CRM systems, you retain full control and ownership, enabling seamless changes or enhancements as needed.
- Long-term cost savings: Contrary to off-the-shelf CRM systems' hidden costs, businesses investing in custom solutions witness an average ROI of £8.71 for every pound spent.
Ready to discuss how a custom CRM system can transform your business? Simply reply to my email and we can schedule a call.
Kind regards,
Andrew Miller
CRM Management Ltd
01759 361 845
crmmanagement.co.uk
I have sent this email to info@learn.odoo.com. If you don't want to get any more emails from us, you can unsubscribe here.
CRM Management Ltd - English House, Railway St, Pocklington, York YO42 2QZ - Company no. 10684534
by "Andrew Miller" <Andrew@crmcompany.co.uk> - 11:34 - 19 Nov 2024 -
The CEO’s essential checklist: Questions every chief executive should be able to answer
Share this email
Brought to you by Andy West (Andy_West@McKinsey.com) and Michael Birshan (Michael_Birshan@McKinsey.com), global leaders of McKinsey’s Strategy & Corporate Finance Practice
New from McKinsey Quarterly
This email contains information about McKinsey’s research, insights, services, or events. By opening our emails or clicking on links, you agree to our use of cookies and web tracking technology. For more information on how we use and protect your information, please review our privacy policy.
You received this email because you subscribed to our McKinsey Quarterly alert list.
Copyright © 2024 | McKinsey & Company, 3 World Trade Center, 175 Greenwich Street, New York, NY 10007
by "McKinsey & Company" <publishing@email.mckinsey.com> - 10:30 - 19 Nov 2024 -
AI-powered observability has already reached 52% adoption. Learn why.
New Relic
Already, more than 50% of software and IT services companies use AI-powered observability to enhance performance, drive resilience, and deliver superior customer experiences.
Find out why in AI And Observability—Unlock Deep Insights and Better Decision-Making. Inside the report, you’ll discover:- Business impacts of AI for observability
- Details of adoption of AI-powered observability by industry
- Where industry insiders think AI is headed
New Relic is bringing you exclusive access to this report because we 100% agree with its central finding: AI is changing observability to sharpen its business impact.
Read Report 
by "New Relic" <emeamarketing@newrelic.com> - 06:03 - 19 Nov 2024 -
What are some of the energy transition’s most demanding challenges?
Only McKinsey Perspectives
Brought to you by Alex Panas, global leader of industries, & Axel Karlsson, global leader of functional practices and growth platforms
Welcome to the latest edition of Only McKinsey Perspectives. We hope you find our insights useful. Let us know what you think at Alex_Panas@McKinsey.com and Axel_Karlsson@McKinsey.com.
—Alex and Axel
•
Early in the transition. Today’s energy system (that is, the production and consumption of energy resources) generates more than 85% of the world’s carbon dioxide emissions, McKinsey Global Institute chair Sven Smit and his coauthors share. Reducing those emissions in just a few decades’ time is a big task: an energy system is a physical entity, and past energy shifts have taken many decades—even centuries—and the current transition remains in its early stages. The deployment of low-emissions technologies is only at about 10% of the levels required by 2050 to reach existing net-zero targets.
•
Challenges to overcome. McKinsey analysis finds that to advance the transition, leaders face 25 physical challenges to address. Of these, focusing on the 12 problems that involve critical interdependencies and gaps in technological performance—which require systemic solutions and are the toughest to solve—could eliminate about half of the energy system’s carbon dioxide emissions. Read the report The hard stuff: Navigating the physical realities of the energy transition to see seven domains of the energy system that must be transformed, and explore our COP29 page to engage with our sustainability experts.
—Edited by Belinda Yu, editor, Atlanta
This email contains information about McKinsey's research, insights, services, or events. By opening our emails or clicking on links, you agree to our use of cookies and web tracking technology. For more information on how we use and protect your information, please review our privacy policy.
You received this email because you subscribed to the Only McKinsey Perspectives newsletter, formerly known as Only McKinsey.
Copyright © 2024 | McKinsey & Company, 3 World Trade Center, 175 Greenwich Street, New York, NY 10007
by "Only McKinsey Perspectives" <publishing@email.mckinsey.com> - 01:59 - 19 Nov 2024 -
Our New Book, Generative AI System Design Interview, Is Now Available
Our New Book, Generative AI System Design Interview, Is Now Available
Forwarded this email? Subscribe here for moreOur new book Generative AI System Design Interview is available on Amazon!
𝐖𝐡𝐚𝐭’𝐬 𝐢𝐧𝐬𝐢𝐝𝐞?
- An insider’s perspective on what interviewers are truly looking for and why.
- A 7-step framework to help you tackle GenAI system design interview questions.
- 10 real-world GenAI system design questions with in-depth solutions.
- 280+ diagrams to demystify complex GenAI systems.
𝐓𝐚𝐛𝐥𝐞 𝐎𝐟 𝐂𝐨𝐧𝐭𝐞𝐧𝐭𝐬
Chapter 1 Introduction and Overview
Chapter 2 Gmail Smart Compose
Chapter 3 Google Translate
Chapter 4 ChatGPT: Personal Assistant Chatbot
Chapter 5 Image Captioning
Chapter 6 Retrieval-Augmented Generation
Chapter 7 Realistic Face Generation
Chapter 8 High-Resolution Image Synthesis
Chapter 9 Text-to-Image Generation
Chapter 10 Personalized Headshot Generation
Chapter 11 Text-to-Video Generation
It has been an absolute pleasure working with Ali Aminian and Hao Sheng. Holding the book in my hands feels incredible. Thank you both for creating such a wonderful piece of work!
If you have any questions, please leave a comment.
Check it out on Amazon now: https://www.amazon.com/Generative-AI-System-Design-Interview/dp/1736049143/
© 2024 ByteByteGo
548 Market Street PMB 72296, San Francisco, CA 94104
Unsubscribe
by "ByteByteGo" <bytebytego@substack.com> - 12:13 - 18 Nov 2024
