Abhinav A.

🚀 Days 10-11 of Headstarter SWE Fellowship: Full steam ahead! 💻 1️⃣ Team sync-up: Finalized our project, MVP features, tech stack, and roadmap. GitHub repo initialized, Trello workspace set up, and Notion page created for easy reference. Excited to see our idea come to life! 2️⃣ Caught the inspiring PathwaysToTech #16 Livestream. Highlights: Julian Alvarez 🚀, CEO of Wisdolia, was the star of the show for me! He emphasized not getting too attached to ideas, noting that over 90% of startup ideas fail. He advises being optimistic about your team's ability to figure things out. He highlighted the importance of taking risks, learning from mistakes, and conducting experiments to validate assumptions and iterate on them.🔬 Additionally, he recommended maintaining health through proper nutrition, sleep, hydration, exercise, and mental wellness.☀ His words truly resonated, especially this gem: "The journey of building something from nothing is as beautiful as it is difficult." I could honestly go on and on about what he said! (it's also nice to hear from a fellow EDM enthusiast working in tech)🎶 Hamza Ali, CEO of Olostep, was also nice to hear from! He is interested in hiring some fellows from Headstarter to work in Olostep! To maximize my chances of getting hired, I'm focusing on: - Honing my #TypeScript, #MongoDB, and #Nodejs skills. - Developing problem-solving skills and customer-centric thinking - Being communicative, humble, and constantly learning While I'm proficient in the #MERN stack, I'm also learning new technologies like #Django, #MySQL, #Vite, and #Tailwind to adapt to my team's needs and fill in the gaps. Always ready to expand my toolkit!💪 3️⃣ Wrapped up by reviewing and providing feedback on Project 2 submissions. Always inspiring to see what the fellows are creating! Going great so far, thank you Yasin Ehsan 🚀and Faizan Ahmed for managing such a big cohort of fellows, and hosting fun events like PathwaysToTech! #HeadstarterSWEFellowship #TechJourney #StartupMindset #ContinuousLearning #AdaptabilityInTech

7

Super excited to share that our Week 3 project for the Headstarter SWE Fellowship is done! This time, we built a customer support chatbot for Silver Spoons, a fine-dining restaurant. The chatbot handles all sorts of things—answering questions about the menu, telling you when the restaurant’s open, and even booking reservations for you. 🍽️🤖 Just drop in the details, and it’s all set! We put this together using: - Next.js and TypeScript for the front-end. - AWS Bedrock, Gemini 1.5 pro and Meta LLaMA 3 70B for the AI smarts. - MUI for the UI. - Firebase for the backend. Our main goal was to make it work smoothly, so the UI isn’t super flashy, but it gets the job done. Big thanks to my teammate, Harrison Akunne and the whole Headstarter team for their support! 🙌 Also big shout out to Bill Zhang and Yasin Ehsan 🚀 for all the support provided. Curious about how it works? Check it out here: https://lnkd.in/ddG3qRrq Can’t wait to tackle what’s next! 🚀 #TechJourney #AI #Chatbot #HeadstarterFellowship #NextJS #AWS #LLaMA3 #Firebase

12

Please check out the article Andrew Dang and I wrote on creating a local chatbot with just 20 lines of code using LLMWare! https://lnkd.in/ecPFDGHu

21

4 Comments

We just made v1 of our app this weekend at the UC Berkeley AI Hackathon! 👨‍🍳 It's still very basic compared to the app we're envisioning, but it’s a good start with one of our most important features (ai question asking) at a stage where we can actually try it out while cooking! Our app name is still TBD. Have any ideas?? - Want to try it out? Join the waitlist: https://lnkd.in/gte54MYD Adhav Rajesh, Saif Alesawy, Nathan Yap buildspace, Cal Hacks, Berkeley SkyDeck

79

14 Comments

📜 A new blog post explaining the fundamentals of RL for LLMs—the perfect place to intuitively understand PPO and RLHF from first principles. 👇👇 Pre-trained LLMs are often compared to Lovecraft's Shoggoth, a monstrous, shape-shifting beast. PPO is the algorithm that brought it under control. Take a look! 🔗 https://lnkd.in/eXqDvrUV

29

1 Comment

It's not often that conferences like AIQCON come around, but sadly, I can’t join this one. Why am I experiencing major FOMO? - Just a one-day conference. - Stellar lineup of speakers, among them: Jerry Liu Emmanuel Ameisen Chip Huyen Shreya Rajpal Lukas Biewald Todd Underwood - Three dynamic tracks covering critical areas: Quality, Regulation and Evaluation, Autonomous Systems & Robotics, LLMs, Foundational Models, and General AI. There’s something for everyone interested in the cutting edge of technology. - High energy organizers like Demetrios Brinkmann that help making MLOps Community such great place. If you're free on June 25 and are passionate about groundbreaking topics in AI, check out AIQCON 2024 https://lnkd.in/e8a4_7kE

10

4 Comments

PSA: if you're on the AICamp newsletter. We're partnered up with the The Linux Foundation to offer 100 FREE tickets (Late bird value at $850) to the upcoming Pytorch Conference. First Come First Serve. Check your email! May the odds be ever in your favor. PS: This is a timed post, and I won't be able to reply until after 5ish - if ya missed out, maybe next time! #LinuxFoundation #pytorch

5

Week 3 at Headstarter Fellowship: Diving Deep into AI Chatbot Development 🤖 https://lnkd.in/gQs-w42f Wrapped up an intense week building a sophisticated AI-powered chatbot. We started with a Next.js foundation, leveraging its API routes for our backend needs. The frontend came to life with Shadcn UI components styled with Tailwind CSS for that sleek, modern look. But the real magic happened when we integrated the AI. We implemented Retrieval Augmented Generation (RAG) using Langchain and Pinecone, which allows our chatbot to pull relevant information from a specified knowledge base (Right now it just has some old information about me). Then came the fun part: we integrated both Groq's Mixtral-8x7B and OpenAI's GPT-4o, building a custom routing mechanism to choose the best model based on query complexity. The chatbot isn't just smart; it's multilingual too! We added support for English, French, and German responses. And to keep things dynamic, we implemented streaming responses - it's pretty cool to see the AI "think" in real-time. While each of these components might seem straightforward on its own, the real challenge was orchestrating everything into a seamless, cohesive application. It's one thing to implement RAG or integrate multiple AI models, but to make them work together smoothly? That's where the real fun begins! There's still room for improvement, of course. We're looking at refining the query complexity determination and optimizing RAG performance. But I'm proud of what we've accomplished in just a week. It's a testament to how quickly we can build complex AI applications with modern tools and frameworks. This project has been a fantastic learning experience, thanks to Bill Zhang for the great help, pushing me to think holistically about AI application architecture. Can't wait to see how we'll apply these skills in future projects! #HeadstarterFellowship #AIEngineering #ChatbotDevelopment #HeadstarterAI Faizan Ahmed

5

🚀 Excited to share my completed Project 2 for the HeadStarter AI SWE Fellowship! 🎉 I’ve developed a Pantry Tracker app using Next.js, Material UI, and Firestore. Key features include: 🔹 Low Stock Alerts: Background color turns red for items with a quantity below 4. 🔹 Search Functionality: Easily filter items by name to quickly find what you need. 🔹 Item Management: Increase, decrease, and delete buttons for straightforward updates to your inventory. 🔹 Upcoming Enhancements: Plan to add a login page for enhanced security and some improvement on the user interface. Due to a challenging health issue that led to my hospitalization, there was a delay in submission. I’m now recovering and back on track! Check out the app and video demo here:  https://lnkd.in/gRsHZvWd Big thanks to the Headstarter team and Bill Zhang for the invaluable support! #HeadStarterAI #SoftwareEngineering #NextJS #MaterialUI #Firebase #AppDevelopment #TechCommunity

4

I wrote some notes this week about the AI Index Report and what it means for everyone. The AI Index Report is an objective overview of the current state of AI written so anyone can understand. Even if you aren't interested in AI, this is an important read to understand how AI will impact you. https://lnkd.in/dQaPYZzN

2

Episode 16 of my "Game-Changing Tech in Product Development" series is now live! This week, we're diving deep into the world of Quantum Computing and how it’s changing the game across industries by shifting from traditional binary systems to qubits. The potential impact is mind-blowing, especially in material science! Key takeaways: 1️⃣ Material Science: Quantum computing’s biggest breakthrough, helping companies like Volkswagen Group process complex data sets to improve materials for electric vehicles. 2️⃣ Automotive: Volkswagen is using quantum algorithms to optimize traffic flow and vehicle design. 3️⃣ Pharmaceuticals: IBM is leveraging quantum to accelerate drug discovery, making personalized medicine a reality. 4️⃣ Finance: Goldman Sachs is exploring quantum for risk management and real-time data optimization. Quantum computing is redefining what’s possible—from tackling complex materials to solving optimization problems at speeds we couldn’t even imagine with traditional computers. Watch the full episode and see how this tech could transform your industry!👇 #QuantumComputing #Volkswagen #IBM #MaterialScience #ProductDevelopment #TechInnovation #EmergingTech #Approlabs #DigitalTransformation

206

49 Comments

Thrilled that mine and Qian Li's paper was just published in the Communications of the ACM! The big challenge we've been working on since grad school is how to make serverless computing work well for stateful applications. This is tricky because it's hard to make long-lived stateful workflows reliable in ephemeral serverless execution environments. The solution, which we talk about in this paper and are building at DBOS, Inc., is using database transactions to make programs durable. The key idea is to store the execution state of a serverless workflow in a database. That way, if the workflow is interrupted, it can automatically recover from where it left off by looking up its execution state in the database. If you know that stateful programs are durable, it becomes a lot easier to make them serverless. You can deploy them to ephemeral executors and leverage their durability to autoscale them and move them between executors as needed. That way, you can build incredibly reliable stateful programs that also benefit from the scalability and convenience of serverless.

110

9 Comments

🚀 Excited to unveil our latest project: Resume Builder! 🎉 Developed as part of the Headstarter AI Software Engineering Fellowship hackathon, this tool is designed to make crafting and managing professional resumes easier than ever. 🎁 What’s Inside: ● Elegant and Professional Design: Choose from a variety of modern templates that showcase your skills and experience effectively. ● Customizable Sections: Tailor your resume to include sections like Experience, Education, Skills, and more. ● User-Friendly Interface: Navigate effortlessly with a simple, intuitive design. ● Multiple Export Options: Save and download your resume in PDF and other formats. 🌟 Huge thanks to my fantastic teammates: Akshat Singh Vikrant Mohan Rishi Upadhyay Your hard work and dedication made this project a success. I’m grateful to the Headstarter AI community for their support and inspiration throughout this journey. Check out our GitHub repository: https://lnkd.in/gCSwTBad to explore the tool, and if you like it, don't forget to star it! 💼✨ 📜 Find the link to website: https://lnkd.in/gD5PzXka #HeadstarterAI #SoftwareEngineering #Hackathon #ResumeBuilder #TechInnovation

29

Introducing Pantry Pie - the Pantry inventory that lets you generate recipes with AI!🍉 Yeah that's me again! (4am post) I'm thrilled (everytime) to share my latest project, built during my second week as a SWE Fellow at Headstarter. I used Next.js and Material UI for the frontend, Firebase for user authentication and Firestore database. I also integrated Gemini 1.5 Flash Model API, which generates recipes based on your pantry staples. (watched a lot of tutorials and read tons of documentation) After 100+ commits, 3 branches for different work, and almost 10+ deployments (due to bugs - it still has some), I'm proud of what I've accomplished. And yes, it's not perfect yet - but as Yasin Ehsan 🚀 said, "5 imperfect projects are better than 1 perfect project". here's the link: https://lnkd.in/dx-xU8Vk GitHub repo: https://lnkd.in/dBsR4rdF I can't believe I learned all this in just 3 days. (I still remember starting web development just one month ago and making my first to-do list app lol) The real software engineer loreee!🗣️ // I ain't good at video editing stuff, so I generated this using AI. It ain't perfect^^ Rate my UI/UX skills!!! #PantryPie #HeadstartAI #SWEFellowship #NextJS #Firebase #RecipeForSuccess" !

62

15 Comments

I wrote about about the state of compute in the world of AI and what that means for individuals. I also included my usual dump of ML resources and updates in a streamlined fashion. I'm changing up the format of my writing this week. Let me know what you think! https://lnkd.in/dS_EXsK5

5

2 Comments

More often than not, I’m here reflecting: What are the implications of making LLMs endlessly larger? My mind visualizes impossible levels of energy consumption, increasing exponentially each day. It feels like a scene from The Matrix, and not a pretty one. Recently, I’ve been researching about RAG (Retrieval-Augmented Generation), and it sounds promising, there are key points showing potential to reshape how we develop and deploy AI. The aspect that stands out the most to me is the potential to shift away from the “Texas approach” (remember my last post?), RAG could let us move towards smaller(!), modular, specialized models that integrate only the data they need, when they need it, rather than scaling endlessly. It’s like my dream of sibling models working together, creating a more targeted approach for better results. Why am I against ginormous models, you might ask again? LLMs can consume as much energy as powering a small city. City. RAG could allow us to significantly cut energy consumption, by eliminating constant retraining and enormous parameter counts. Have I mentioned I dislike the massive cash burning? Yes, is not my money. Yes, who cares about the overlords money. But honestly, I just hate waste. Did you know the cost of training foundation models have been doubling every 6-10 months? Doubling. Read again. RAG offers the potential to invest once in a base model and then build on it through knowledge retrieval, rather than constant retraining. This approach could make AI access more affordable, opening doors for smaller organizations and underserved communities. Of course, the cynic in me asks: are we simply shifting the computational burden elsewhere? RAG could reduce training costs, but it would introduce database management and retrieval overhead. My cent and a half? What do you think is more costly, monetarily and environmentally, that overhead, or energy consumption? I think you all know my answer. But let’s not forget: RAG doesn’t rely solely on LLM parameters; it retrieves data in real-time from external sources. Meaning APIs, websites, organizational databases, and beyond, and this retrieval approach possess risks and requires robust data security. However, again, the cost will still likely so much lower than today’s soaring energy bills and carbon footprint. So, the question for the 15$ in my wallet is: will AI’s future depend on more power or will be ruled by better processes? My other half cent? A leaner, privacy-conscious, eco-friendlier AI isn’t just a possibility; it should be our goal. All of us. #AI #SustainableTech #EnergyEfficientAI #DataPrivacy

18

6 Comments

🚀 Day 2 of the Headstarter AI Software Engineering Fellowship was thought-provoking with Faizan Ahmed and Jack O'Brien! Here are some of my favorite insights: 🔹 "Distribution eats product for breakfast" – Emphasize growth product features and leverage short-form content on platforms like TikTok and Instagram to capture attention. 🔹 Building relationships with startups: Platforms like Y Combinator can help find promising startups that may resonate with our education, interests or experience. Focus on building rapport, testing their products, and offering constructive feedback. 📌 Tips to build a project from scratch: Waitlists are great: Create a landing page, test engagement with a basic MVP, or provide free credits to gather emails from interested users. Blogging and meetings: Write valuable blog posts and network in real life. 💡 Jack O’Brien’s tips: Side projects: Critical for career growth. His personal examples, like a presentation platform for startups and an environmental calculator, were inspiring and motivating. Don’t obsess over the “idea”; just choose a decent one and start practicing. Getting first users: Work on projects that excite you, involve friends, and share on as many platforms as possible. Using AI to code: Embrace tools like Cursor AI and ChatGPT, but ensure you understand what your code is doing. The overall takeaway was to get in the reps! This event was insightful and inspiring, and I’ll take action on these tips. Thank you, Faizan and Jack, for this great event! Excited for what's next in the program! #HeadstarterAI

4

AI at Meta 🤝 Roboflow: Open source computer vision New post on the Meta AI blog: how Segment Anything has saved over 74 𝐲𝐞𝐚𝐫𝐬 (and counting) for hundreds of thousands of users curating computer vision datasets on Roboflow. Amazingly, even this under estimates SAM's measured impact by an order of magnitude, i.e. all our open source work can be self hosted: using SAM in Workflows for chaining together models in inference, running SAM on edge devices, and with autodistill for local distillation. Link in comments ↓ #opensource #computervision #segmentanything #AI

278

13 Comments

Data Science Conference Data Science Academy GenAI Summit 🚀 Looking Forward to Exploring Multi-Agent Systems with Wyzard AI! 🤖✨ Excited to announce that we’ll be joining an insightful session on Multi-Agent Systems with Wyzard AI on 19th December! 🎉 Can’t wait to dive deep into how collaborative intelligence and autonomous systems are revolutionizing the AI landscape. Stay tuned for updates and key takeaways! 🌐💡 #MultiAgentSystems #ArtificialIntelligence #GenAI #WyzardAI #AIFuture #AIInnovation #TechLeadership #AIApplications #TechTrends #AICommunity

2