Arif Zaman

🕊️ Peace is my long standing software automation project where I've engineered my heart into code. Video: https://lnkd.in/ggfs3amK (45 mins) - watch at 1.5x speed, will not disappoint 😊 Features (surprisingly good): 1. 🖼️ See a diagram of the environment that would be created, before it exists. 2. 👁️ See a diagram of the environment that actually exists. 3. 🎨 See what would happen at every step of the process before you press go. 4. 🖌️ See what's actually going on as it is happening. 5. ⚡ Interrupt and resume automation safely. 6. 💻 Beautiful command line and web interfaces generated for you. 7. 📄 "Like git status, diff, commit, and push, but for servers" Links: * Slides: https://lnkd.in/gxetkv-T * Website: https://peace.mk/ * Code: https://lnkd.in/g8DM2UN2 * Diagram: https://lnkd.in/gZQWVK5y To engineer with empathy, whether verbal, visual, or vocal, refine your voice, connect, and communicate with clarity. Let it reach you.

22

Reliability is still one of the main issues plaguing AI agents in the wild. Sohil and I know this because our coding agent Tusk (YC W24) pushes code to production apps for customers with mature, complex codebases. As such, we're always looking for step-level improvements in code generation quality. Proud to show off some new agent improvements that have achieved this leap for our AI agent: 🧑‍💻 Tusk now refers to a semantic graph when navigating your codebase for context. This allows it to more reliably find files and symbols that are relevant to your engineering task. 🧠 Tusk now uses past merged PRs as long term memory to solve future tasks. This replicates how humans recall completing similar tasks in the past when given a new task. Check out the video below showing how these improvements assisted in solving a user-reported bug with our marketing website––all from a Loom video. Thanks Ilya Subkhankulov for the bug report :)

78

2 Comments

My Facebook and Instagram accounts got hacked. Someone created a fake account on Instagram, linked to mine, violated the rules, and as a result all my Meta-related accounts got suspended for good. 16 years of being a good Facebook citizen, not a single suspension, not a single warning, over a thousand of contacts - family, colleagues, friends, partners, clients - conversations, photos, videos, memories... some of it is in the backup, most are not. All gone thanks to what looks like CSRF attack that happened after clicking an ad that was running on Facebook itself. 2FA did not help - if you are already logged in, 2FA is not required to create another account that can be associated with any email. What Meta does about it? Absolutely nothing. No way to appeal, no way to contact their support, not a single way to restore 16 years of my online life. No solution from their help works. My wife got Meta Verified badge and was able to contact their support on my behalf but they just sent us the same "self-service" links that did not work. We showed them screenshots of failures, we sent them government IDs. They apologized for not being able to help and closed the case. I've heard many horror stories about how Meta treats its users but now I am living through one. Is legal direction the only option to get their attention?

8

11 Comments

How do you build authentic, meaningful connections in a remote-first, global tech world as a software engineer? I’ve been working remotely for a Japanese tech company while living in Bangladesh for the past 4 years, and honestly, I sometimes find myself wondering—where do I truly fit in the professional network landscape? I’m not deeply connected to the local tech community in Bangladesh, nor to the tech scene in Japan, where my company is based. Networking sometimes feels like a bit of a gray area for me. And how much of a meaningful network is realistically possible, being just a software engineer? How much of it is truly necessary? Would love to hear your thoughts. How do you navigate networking in a setup like this?

35

11 Comments

One of my highlights of this week was definitely the Pathways to Tech Event thanks to Headstarter AI. With a unique lineup of speakers from different perspectives of the SWE field, Nabeel Alamgir🦄 (Lunchbox CEO), Shariar Kabir (Ruby CEO), Dustin Beadle (Lattice Director), and more shared such valuable wisdom with all who were in the Zoom. Nabeel's feats were quite impressive, as he faced rejections from over 72 investors before having any support/rapport. Hearing about his consistent resilience and determination gave me courage. He also talked about his own journey, and that we have to stay true to ourselves, but also be able to build trustworthy and respectful relationships with others. He also put heavy emphasis on learning from our failures being the key to our success. One of the things he said that stuck with me was to not oversell yourself, and be honest. Shariar also talked about failures and their evolution into learning opportunities. It's also key to have a diverse background, as Shariar was Ex-Salesforce, and Miguel Acero, also one of the founders, was an Ex-Google exployee. They also both touched down upon hackathons and side projects, and that they can be vital stepping stones for your own personal growth. Finally, continue to have persistence. Dustin Beadle caught my eye the most, seeing as he is a Ex-Director at AMD. He talked about the importance of cross-functional skills (ex. adaptability in multiple roles), and how comp sci students should learn to be able to communicate effectively throughout different roles. Similar to what Yasin Ehsan 🚀 has said earlier during his time at Capital One, staying curious is essential in the fast-paced technological world. Overall, to prepare for a possibility of maximizing my time and increasing my chances, I will continue to build a strong network with the other fellows in Headstarter, continue to push myself to adapt to new languages frequently, and embrace the challenges of front-end, back-end, etc.

17

1 Comment

3 months of study and I got a new certification

36

1 Comment

A few use cases where AI can solve real problems: I was impressed by Zee Hussain's post, where he said, "Focus on real pain, and create pain killers, not vitamins." I conducted some experiments to explore how AI can help create pain killers. You'll be amazed to learn how Radiologists can use medical imaging analysis to diagnose and treat patients more quickly. --------------------------------------------------------------------- #ai #genai #aipm #pms #pm #productmanagement #productmanagers #ResponsibleAI #UnbiasAI #DataScience #MachineLearning #MLOps #openai #llm #slm #ollama #llama #AIOps #kubernetes #cicd #pipeline #python #devops #devsecops #data

8

1 Comment

🌍 Moving to Germany from Bangladesh involved a lot of shopping, and I wanted to avoid spending a fortune on products which is made in Bangladesh. Instead, I decided to buy a variety of reasonably priced items from home. Moving abroad can be a real challenge, especially with all the searching and carrying heavy shopping bags, which can be quite painful during long walks. To make my shopping experience hassle-free, I add a simple extension for my shopping bag which also found in the market. Not only did it make my life easier, but it also brought some laughs during my time in the market! I hope this idea can help others in similar situations. 😊 #MovingAbroad #ProblemSolving #ShoppingSolutions #BangladeshToGermany #CreativeSolutions #DeveloperLife

12

A data-driven approach to driving wound healing outcomes Marcella Gomez, University of California, Santa Cruz https://lnkd.in/gAkWZYJT Meet: https://lnkd.in/g4AAWkcK YouTube Stream: https://lnkd.in/gwtqaTmf Talk details: https://lnkd.in/gWn3B3cU Subscribe for talk announcements: https://lnkd.in/g5ciuNuX Abstract: Precision medicine requires an ability to predict the response of an individual to a prescribed treatment regimen a priori. Thus, advancement in the field is challenged by a lack of predictive models and, arguably, a lack of time-series information for a highly dynamic system. Here, we discuss work in wound healing, for accelerating wound closure. We argue that timing of treatments is as critical to consider as the choice of drug or therapy. Due to system size and complexity, data-driven methods need to be explored to develop multi-dimensional quantifiable indicators tracking systemic changes. In this work I discuss how bioelectronic devices enhanced with deep learning can help facilitate real-time sensing and actuation for automated decisions in treatment for wound healing and preliminary work in transcriptomic based classification of wound states. Bio: Marcella M. Gomez is an associate professor at UC Santa Cruz in the department of Applied Mathematics and Associate Dean for Diversity, Equity, and Inclusion for Baskin Engineering. She received her PhD from Caltech in 2015 and a B.S. from UC Berkeley in 2009; both degrees in Mechanical Engineering. Her research interests are in the broad field of bio-control leveraging methods in machine learning and control theory. Applications range from controlling single-cell response to driving complex systems such as wound healing. #modeling #simulation #medicine #healing #trauma #predictivemodeling #datascience #data #woundhealing #biology #bioengineering #systemanalysis #sensing #omics

3

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

Check out our recent work on “Maximizing GPU Utilization through Apache YuniKorn Preemption” which will be presented today at Community Over Code 2024 by Weiwei Yang and Junyan Ling : https://lnkd.in/gn4Ch_qq Efficient GPU utilization and cost-to-serve optimization have become critical for running ML inference jobs on GPU clusters due to the limited and expensive nature of GPU resources, as well as the variation in workloads, job frequencies, and sensitivity to execution time. In this talk, they will share practical insights and best practices for tackling these challenges. They’ll demonstrate how Apache YuniKorn’s preemption feature significantly improves GPU utilization, reduces scheduling delays, and increases inference throughput.

72

3 Comments

Generative AI (Gen AI) is currently a hot trend due to its widespread adoption in all phases of the software life cycle, regardless of the business type. As we focus on quality criteria at each stage, it's crucial to delve into how Gen AI is validated with the highest quality. While tools and frameworks play a role, adhering to strong principles is paramount. Here are some key aspects to consider for ensuring the quality of Gen AI framework: - **Relevance**: Ensuring that generated data is relevant to the context is essential. Irrelevant data, no matter how vast, serves no purpose and is not valuable. - **Accuracy**: After relevance, the accuracy of the generated data is critical. Precision in data generation is vital as inaccurate data can lead to erroneous outcomes. - **Sensitivity**: Sensitivity plays a significant role, especially with specific datasets and considering factors like age, gender, and geography to comply with policies. - **Security**: Securing data with appropriate permissions and privileges is a top priority. Protecting sensitive information such as personal banking, financial, and health data is crucial. - **Scalability**: In addition to functionality, scalability is key. It is important to assess if the framework can scale effectively to meet the needs of enterprise applications. Do you have any thoughts to add to enhance the quality of Gen AI frameworks? Share your insights and contribute to the discussion! #GenAI #Quality

23

Amazon-Meta-Microsoft Developer Productivity dinner. Last night I had the privilege of getting together with Brian Houck (Senior Principal at Microsoft) and Karim Nakad (Senior Staff at Meta) and we shared a leisurely two and half hour meal talking about DPE in the industry, and life in tech. We exchanged interesting ideas, picked each other’s brains, and found an amazing amount of common ground between us, despite working in entirely different companies. One thing I really appreciate about the Seattle area is that I’m surrounded by tech talent, not just within my own company, but industry-wide. This means I can easily grab coffee with friends from Meta, Microsoft, Uber, Google, Oracle, Snowflake, just to mention a few. I think this is critical and I’ll tell you why. I’ve written about “bubbles” in the past (https://lnkd.in/ggV7EbMk) so I’ll lazily quote from there: “My entire 27 year career in software engineering has been at three out of the Big Tech companies: Amazon (12yrs), Microsoft (11yrs) and Google (4yrs). And I have an embarrassing admission. I lived in bubbles. Initially, I was blissfully unaware I did: it wasn’t deliberate, it was simply naive. Today, I still live in bubbles, but at least I’m aware and I take steps to alleviate the negative effects. Because living in bubbles is unavoidable. I’m using the term ‘bubble’ as in ‘assuming that most people in the world live and think like you, because most people around you live and think like you’ ” I strongly believe it is critical to look beyond your company. What are other people in the industry doing? I think about Developer Productivity a certain way - but what do other smart people at Meta, Microsoft, think? All in all, I think there’s more common ground than not. We all care deeply about Developer Focus - too many things these days (chats, emails) break focus. We are trying to understand the impact of GenAI on engprod. We want to understand and quantify Developer Productivity. And we obsess about how to help our tens of thousands of developers be happier and more efficient in what they do. #amazon #meta #microsoft #engineeringproductivity #dpe

495

23 Comments

Update: meeting link added in comment. Wow! I didn't expect so many of you would join a call at an instant notice with literally no agenda. And I loved the questions and interactions! Thank you for joining. This also is a means for me to know the insights of job-seeker world, first hand. As promised in today's sessions, I will try to conduct one more tomorrow and will start with resume reviews this time. Will share the link shortly. #jobs #faang #interviews #mentor #hiring #resume #review

9

11 Comments

⏳ Sometime back I interviewed ✍ a Site Reliability Engineer (SRE) for my team, and the difference between a good SRE and a great one became evident through a simple question about curl and how it works. While a typical answer would cover the basics of how curl operates, the response I received showcased the depth of expertise a great SRE possesses. - Resolving the URL using internal system calls like getaddrinfo(). - Establishing a connection with the server by creating a network socket with socket() and initiating a TCP connection using connect(). - Sending an HTTP request through the socket via send(). - Receiving and displaying the server's response by receiving data with recv(), utilizing poll()/select() for non-blocking I/O, configuring socket options with setsockopt(), and closing the socket connection with close(). This level of detail and understanding of the internal workings of tools like curl truly sets exceptional SREs apart. It's this kind of knowledge and expertise that makes a significant difference in ensuring the reliability and performance of systems. #SRE #Curl #SystemCalls #Expertise

36

2 Comments

Conflict. It is unavoidable experience when raising children. Similarly, in a work environment, if we truly value diverse thoughts, cultures, and approaches, conflict will eventually arise. It is natural to feel inclined to take a side and quickly resolve the conflict or to avoid it altogether. However, neither of these approaches are ideal in the long term. Some time ago, I wrote about the leadership lessons we can learn from our kids and how quickly they grow. I have recently been pondering on this topic as I have had to learn how to negotiate effectively with my kids. "Do it because I say so" is never a good idea. In the parenting realm as well as the leadership context. Fortunately, there are plenty of sources we can learn from. One of them is a book written by a former FBI hostage negotiator. Yes, you heard it right. We can implement the techniques used in negotiations with kidnappers and apply them to our families and leadership at work. I recently started implementing more intentionally some of the lessons I learned after reading the book "Never split the deference" by Christopher Voss. Chris is a former FBI hostage negotiator and the founder and CEO of the The Black Swan Group, a consultancy that specializes in negotiating and conflict resolution. He is a great speaker and the author of several books, including "Never Split the Difference". Below I share the main tools that I find the most useful. 📌 Tactical Empathy The ability to use empathy in a strategic way, to get the other person to do what you want. My first talent in Gallup assessment is a Relator, hence it's more natural for me. But this a skill and can be trained. 📌 Labelling and mirroring The best way to help someone think is to encourage them to reflect on what they are saying. The best interviewers like Andrew Huberman or Tim Ferriss know about it! Pay attention the next time you listen to your favourite podcast. Examples of labelling language: - it seems like... - it sounds like... - it looks like... 📌 Setting the right tone Use the late-night FM DJ voice: slow, deep, and calm. Humans naturally copy the energy of the people they're around. It helps to deescalate the conflict. If people cannot provoke you to rise your voice - you already possess the superpower! 📌 Calibrated questions Examples:  - What is the biggest challenge you face?  - What about this is important to you?  - How can I help to make this better for us?  - How would you like me to proceed? - What is it that brought us into this situation - How can we solve this problem?  - What’s the objective? What are we trying to accomplish here?  - How am I supposed to do that? --- Link to my notes for that book and some anecdotes below 👇 P.S. The photo I took in Iceland illustrates what is left after the classic negotiations, where each side is fighting to win. The place consumed by fire... #book #leadership #parenting

9

2 Comments

Let us shift our perspective on AI and view it as a valuable asset rather than a threat to humanity. The advancements in technology and intelligence have significantly enhanced our lives, providing us with comfort and convenience. In the past, our ancestors faced immense challenges due to the lack of developed tools and technology. However, with the aid of technology, we have made remarkable discoveries about the universe, the world around us, and even set foot on the Moon. Moreover, technology enables us to communicate effortlessly with friends and loved ones, while also facilitating continuous learning and exploration. These are just a few examples of how technology has improved the quality of human life. Admittedly, AI will automate and replace certain jobs, relieving individuals from repetitive tasks. However, this will ultimately prove beneficial as it allows people to focus on more significant endeavors that bring greater happiness. As some jobs become obsolete, new opportunities will arise, shaping a more modern society. AI will not be able to replace human interaction, emotion, and relationships, it will only make them stronger. Let us appreciate the value that AI brings and recognize the positive impact it will have on the future of our society. 🤖🤍 #AI #GenerativeAI #modernsociety #technology #jobs #discoveries #perspective

1