Sara Khodeir

One thing I’ve learned in my first 10 months at Meta… Don’t ask a senior eng what to do or how to do something. Instead, explore different approaches and ask the senior eng which they would recommend. For example, when devising the tech plan for a new feature that requires storing and retrieving new data, don’t simply ask what data solution you should use. Instead, scour internal wikis, existing code patterns, etc, to learn about the data storage solutions available. Then list out the pros and cons specific to your use case in terms of latency cost, capacity usage, reliability, implementation complexity, etc. Sending that breakdown to the senior eng (along with your own proposal for what you think the best choice is) will not only make their lives easier, but prove to them that you’re capable of working at the next level. Of course, there are exceptions. For example, if there’s a quick question that a senior engineer could answer in 30 seconds, but would take you 2 hours to figure out on your own, it makes sense to ask. What do others think? What types of questions should you ask senior eng and how should you decide whether to ask a question or figure it out on your own? #softwareengineer

1,329

138 Comments

My first week as a SWE Fellow at Headstarter AI has been an incredible journey, filled with learning, growth, and valuable insights. One of the highlights was attending the "Our Pathways to Tech" event, hosted by Yasin Ehsan 🚀, Faizan Ahmed, and the Headstarter team. The event featured amazing speakers like Nabeel Alamgir🦄, Shariar Kabir, Miguel Acero, and Dustin Beadle, who shared their wisdom on thriving in the tech industry. Here are some of the key takeaways that resonated with me: 1. Nabeel Alamgir's advice about taking ownership of our decisions and viewing rejections as opportunities to improve was particularly impactful. This mindset is essential for both professional and personal growth. 2. The importance of reaching out and building connections within the industry cannot be overstated. Networking and fostering relationships are crucial for career development and finding new opportunities. 3. Being part of Headstarter AI has reinforced the value of maintaining high standards and holding oneself accountable. This experience has taught me the importance of hard work, effective communication, and the joy of being part of a diverse community of developers. 4. The potential for job opportunities from companies like Ruby, founded by Shariar Kabir and Miguel Acero, was a motivating factor. It encourages me to work harder and strive for excellence, hoping to secure similar opportunities. I am excited about the challenges and learning experiences that lie ahead in the coming weeks. Ready to take on the hits and grow from them! #TechIndustry #CareerGrowth #Networking #HardWork #SWEFellowship #HeadstarterAI #Inspiration #Motivation

5

Impressive! It can tackle math problems, aid in interview preparation, provide real-time translations, sing songs, scan and explain surroundings, assist visually impaired individuals in understanding their environment, and much more. OpenAI GPT-4o is transforming human-computer interaction! It seamlessly processes text, audio, and images with lightning-fast responses. With enhanced language comprehension and multi-modal understanding, it's a game-changer for natural interactions. #ai #gpt4o #llm #datascience #machinelearning #deeplearning #openai #genai

4

Over the last 1 year, I've dived deep into the codebases of major projects like TensorFlow, MongoDB, Chromium, OpenSSL, LLVM, GCC, Kafka, Qt, CMake, and currently the Linux kernel. Transitioning from data structures and Development using Javascript based Tech Stacks to exploring open source C++ projects felt like stepping into an entirely new world. What I discovered were common practices and patterns that define production-level C++ code across these projects. Curious to know what they are? I've compiled my insights into a PDF—perfect for anyone looking to explore the Open Source Projects and contribute to the same. #C++ #OpenSource #SoftwareDevelopment #Programming

86

8 Comments

I'm interested to know what most people might think of this? ❔ So I just read from a post by Al Jazeera Media Network on Instagram that said "Gen-Z and Millennials are shifting towards old-school phones", primarily because of how addictive social apps are these days and how they spent most of their time viewing screens. It also explained how, as a result of this, companies are emerging that are solely focused on developing these "old-school" phones to cater this and I think they actually might capture some traction. But what if they don't need to be old-school? What if we've had these AI-Powered Mini-phones, that rely on voice commands and advanced scheduling and automation features which would enable all the "productive tasks" and yet keep all the dirt away? Let me know what you think of this concept? An AI-Generated concept is attached as well. #research #ai #smartphones #futureoftechnology

17

6 Comments

While working with large language models (LLMs) for chatbots, I've found the invoke function in LangChain quite irritating (this is my subjective opinion xD). It processes queries and displays the complete response only after it's fully generated, making the chatbot seem unresponsive. Found the stream function in LangChain to be very helpful — an absolute game-changer! It streams output in chunks, showing intermediate progress just like ChatGPT. This makes interactions feel much more responsive and engaging. The stream function processes and yields each chunk as soon as it's available. This can range from simple tasks, like emitting tokens produced by an LLM, to more complex ones, like streaming parts of JSON results before the entire JSON is complete. This strategy is important because LLMs can take several seconds to generate a complete response, which is much slower than the ~200-300 ms threshold for a responsive application. Attached video is demonstrating the Llama 3.1 model running locally, showing how the streaming function works and can improve chatbot interactions. Check it out! Feel free to reach out if you have any questions or want to know more about using LLMs in your applications. Would love to help. #langchain #llms #llama3.1 #chatbots #AI

14

I resigned from my dream of working in Amazon for a 6 months’ vacation. In the end it will be an intense 11 years and 8 months experience, started as Individual Contributor followed by 8 years as Software Development Manager, in 5 organisations in both SDO and AWS. Adhered to Constantin Brancusi life’s motto “Work like a slave; command like a king; create like a god”, being twice promoted vertically and once horizontally, while making it to top 1% tenure in Romania and 1.3% global. Fully embraced all the LPs but loved Learn and Be Curious, Ownership, Develop the Best, and Deliver Results hence became Amazon global Samurai, European Samurai, Quality Samurai Program Owner, Romanian Samurai, and Engineering Excellence Program Contributor. “Samurai-ng” offered the chance to review tens of tech designs, contribute and simultaneously learn from brilliant engineers and (under Raluca Buhaianu guidance) created, delivered, and collaborated at training material for tens of thousands of Amazonians around the world, basically supporting the ramping up of every newcomer engineer or engineering manager. As a manager learned first-hand from the wealth of experience of Executives like Lucian Vornicu, Dennis Suppe, Blair Hotchkies, Wyatt Camp, Reena Agarwal, Claire Ngo, Noopur Vyas, stepping out of the comfort zone with every major release. Reporting directly to you and having weekly 1:1s put my growth on steroids, much obliged! Had the chance to leading charters up to multiple teams and to becoming a second line manager thanks to Raluca Buhaianu, Ciprian Răzvan Hodorogea, Dragos Cojocaru, group owners tailoring Company-wide solutions around talents first. Forever grateful to you, major shareholders of my story! When need and opportunity occurred, learned from formal and informal mentors like Garry Gross, Razvan Udrea, Mihai Diac, Bogdan Ghidireac, Cristi Balcanu, Florin Manolache. You impacted my personal growth and my teams' development, thank you! Received so much but also able to leave the Amazon Development Center Romania world a little better, doubling down on Developing the Best, currently ranking first for vertical and horizontal promotions for software engineers (SDE, SDET, FE Eng, AS, WebDev, Support Eng, SysDev, BIE, TPM, SDM), with the exception of Data Engineers (no.2 if including DEs too, after Andrei-Liviu Dumbrava, a role model Sr. SDM I look up to). Amazonians are knowledgeable, gritty, and resourceful, being a complete privilege partnering on projects ranging from builder tools to text and image extraction and processing, or to Payments, Privacy and Customer Behaviour platforms, making or saving BN$ in CAPEX, GMS or fines. Thank you to each and every engineer I worked with, I was honoured to collaborate with you! Wishing you all the best, ex-colleagues and now friends, yet hoping not to be final and perpetual, the “short word with teeth sharp to bite through the string that ties past to the future” (Steinbeck): good-by!

311

19 Comments

Principal Engineer Staff Engineer Distinguished Engineer (my favourite) I’m going to say it and it’s likely a little rhetorical: I dislike all these terms immensely. Please can someone give me the canonical description for what they are supposed to mean? (I know I can use a search engine or AI to do it… I want to know what people think they mean). The industry I work in is in danger of disappearing up its own backside. What’s wrong with: Software engineer Why do we need to differentiate? The only reason I can see is signposting to people with differing levels of experience or different sets of responsibilities. If we need to do that then, I’m not sure our teams are set up correctly in the first place. #software #softwareengineering #softwaredevelopment #productdevelopment

98

105 Comments

Reflecting on my first week as a SWE Fellow at Headstarter, I'm amazed at the wealth of experience I've gained. My heartfelt thanks to Yasin Ehsan 🚀, Faizan Ahmed, and the entire Headstarter AI team for their guidance. This week has been transformative, focusing on communication, project development, and technical interview preparation. Here are my key takeaways: 1. Portfolio Website Development 💻: Successfully created and deployed a portfolio showcasing my software engineering skills and projects. This hands-on experience significantly enhanced my web deployment knowledge especially.(You can check out the website at: marcossanabria.com) 2. AI-Powered Mock Technical Interview 🗒: Engaged in a challenging mock interview with an AI personality, pushing me to devise efficient solutions under pressure and refine my problem-solving skills. 3. Inspirational Speaker Series 🤝: Attended an enlightening session featuring industry leaders Shariar Kabir, Miguel Acero, Dustin Beadle, and Nabeel Alamgir🦄. Their insights underscored the critical balance between technical prowess and soft skills, emphasizing the importance of confidence and proactive goal-pursuit in career development. This experience has not only expanded my professional network but also deepened my understanding of the multifaceted nature of success in tech. I'm energized and eager for the coming weeks, anticipating further growth and learning opportunities. #headstarterAI #softwareengineering #SWE #careerdevelopment #fellowship #continuouslearning

8

2 Comments

Hello Everyone, Welcome to Part 2 of our series on Building a Cache Iteratively. In this instalment, we'll delve into the concept of introducing threads into the equation, understanding the challenges of concurrent access, and devising a solution within a data structure. We'll tackle these issues in the most basic yet effective manner, gradually building a reliable system. You can find the detailed discussion in the following Medium article: https://lnkd.in/gsKYTyzN For those interested in exploring the code base, it's available on GitHub: https://lnkd.in/gp-dTVdU Feel free to dive in, explore, and share your thoughts! #LRUCache #Cache #LLD

14

The policies being driven forward by Thierry Breton around the EU AI Act have been utmost damaging for the European tech industry, and they need to be revised as soon as possible. We need to better balance driving European innovation through data collection and modeling with conflicting needs such as data protection and safety. We need to ensure we can turn these developments into businesses that can be the pillar of the next generation of companies that provide productivity growth in Europe. We have no other choice - either we do this here, or we will have to take over products developed in the US and in China, as it already happened with search engines, cell phones, and social networks. We wouldn't want to let this happen with AI applications in robotics or digital assistants. #euacc #euaiact #aiact

9

I had written a blog post on the design of Youtube's Homepage Recommendations a while ago which is my notes and commentary on the original design of that system as is outlined in their paper from 2016. I found the paper to be a very good resource on production ML systems since it's focus is on the overall system, product KPIs and the design tradeoffs over simply introducing a new approach. If you're interested in Information Retrieval and Deep Learning for IR and recommendations (and have never read the paper), you might find my post useful since it also summarises a bunch of critical citations. Open to feedback and suggestions on how I can improve my writing.

18

🎉 Excited to share my first, first-author paper, "EHRMamba: Towards Generalizable and Scalable Foundation Models for Electronic Health Records," co-authored with my amazing colleagues Mahshid Alinoori, Arash Afkanpour, and Amrit Krishnan at the Vector Institute! 🚀 🏥 Electronic Health Records (EHR) capture detailed patient medical histories and hold immense potential for revolutionizing personalized medicine and clinical decision-making. However, deploying EHR models in real-world settings is challenging due to high computational demands and limited hospital resources. Traditional transformer models, with their quadratic computational costs, are impractical for processing extensive EHR sequences. Additionally, maintaining separate downstream models for each clinical task complicates deployability. 🐍 To overcome these limitations, we introduce EHRMamba, a robust foundation model built on the Mamba architecture. EHRMamba can process sequences up to four times longer than previous models due to its linear computational cost. We also introduce a novel approach to Multitask Prompted Finetuning (MPF) for EHR data, which enables EHRMamba to simultaneously learn multiple clinical tasks in a single finetuning phase, significantly enhancing deployment and cross-task generalization. Our evaluations on the MIMIC-IV dataset demonstrate that EHRMamba advances state-of-the-art performance across 6 major clinical tasks and excels in EHR forecasting, marking a significant leap forward in the field. 👨‍🍳 Alongside EHRMamba, we open-source Odyssey, a toolkit designed to support the development and deployment of EHR foundation models, with an emphasis on data standardization and interpretability. 🔥 Discover more about our work and explore EHRMamba and Odyssey: https://lnkd.in/eDYspsuA #EHRMamba #Odyssey #ElectronicHealthRecords #Healthcare #MachineLearning #AIInHealthcare #VectorInstitute

166

13 Comments

🚀 How to Approach a System Design Interview: Amazon Prime Video Example When tackling a system design problem, especially for a complex platform like Amazon Prime Video, it's crucial to approach it strategically. Here's how I break it down: Get clarity on the requirements, both functional and non-functional. Focus on the user flows, define the scale, detail out the components, improve scalability, and wrap up confidently. Pro Tip for Interviews: Don’t rush into drawing diagrams! Ask clarifying questions, understand the problem fully, and then build the system piece by piece. Communicate your thought process clearly, and focus on scalability, reliability, and cost-efficiency. Remember, the trick is to think aloud—explain why you're making certain decisions. The interviewer wants to understand your reasoning as much as the final solution! Good luck to all those preparing for system design interviews! You've got this!💪 Want to master system design interviews and land those $500k+ TPM roles? Make sure you're fully prepared with the Kracd.com interview prep course. Sign up today and transform your career in just 30 days. #SystemDesign #AmazonPrimeVideo #TechInterview #InterviewTips #Scalability #CDN #TechCareers #SoftwareEngineering #Microservices #TechLeadership

9

Another significant development in the field of AI is the emergence of GPT-4o, as highlighted in this link: https://lnkd.in/dy6Ra2ud. Teaching mathematics https://lnkd.in/dCnDjrRv

1

So recently I explored Ollama and its pretty amazing ! What is Ollama? We might not have access to OpenAI APIs or a cloud environment like AWS to run LLMs. What if there was a simple and easy way to run small LLMs locally with zero hassle + not getting charged where we can start to just play with the LLMs and build POCs. Enter Ollama, a platform that makes local development with open-source large language models a breeze. With Ollama, everything you need to run an LLM—model weights and all of the config—is packaged into a single Modelfile. Think Docker for LLMs. And its extremely easy. You just have to go download it from the official website - https://ollama.com/ , install it and in the terminal run the command like - ollama run gemma to download the opensource LLM. And following is an example of how you can use it. Hurray ! You are able to run your own LLM. That being said obviously this is not for production level systems and just for us to start :) Source - https://lnkd.in/ddCbkbWM

17

I recently had the privilege of attending an enlightening tech leaders panel hosted by Headstarter AI. Sincere thanks to Yasin Ehsan 🚀and Faizan Ahmed for this invaluable opportunity. Below are some pivotal insights from esteemed industry leaders: Dustin Beadle (Director at Lattice, Former Director at AMD): • Highlighted the significance of cross-functional skills, effective communication, and a positive attitude. • Encouraged acknowledging imposter syndrome, but not allowing it to impede progress. • Advocated for asking thoughtful questions and practicing active listening. Logan Havern (CEO of Datalogz, Raised $12M): • Stressed the importance of taking initiative and independently solving problems. • Advised mastering self-promotion and customer support skills. • Shared his perseverance story of pitching to 93 ventures in 4 months. Nabeel Alamgir🦄 (CEO of Lunchbox, Raised $100M in VC Funding): • Emphasized learning from mistakes and showcasing resilience. • Urged authenticity; neither overselling nor underselling oneself. • Encouraged striving to become a respected leader. Shariar Kabir & Miguel Acero (Ruby Card (YC W23): • Promoted embracing failure as a learning opportunity. • Highlighted that persistence can lead to success. • Suggested leveraging hackathons and side projects as potential career catalysts. Key takeaways: • Blend technical skills with effective communication and self-promotion. • Initiate conversations with genuine compliments and thoughtful questions. #CareerinTech #LeadershipInsights #HeadstarterAI #ProfessionalDevelopment #TechIndustry #Headstarter #SWE #Fellowship

10

New partnership between Stack Overflow and Open AI. I will be super interested in seeing how this impacts the short and long term accuracy of LLMs in the future. In this video Theo shares a very thought-provoking picture of what this could mean for LLMs in the future. https://lnkd.in/gph_wGPs

1

2 Comments