Adrian P.

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

Santana Row shopping center in San Jose has implemented a ticketless parking system that eliminates the need for drivers to take a ticket upon entry and pay at a pay station upon exit. Instead, drivers can simply scan a QR code (see photo) posted throughout the garage using their phones, enter their license plate number and credit card information on the website that pops up, and a camera will read their license plate and open the gate as they exit. While this system may seem convenient and efficient, I think there is a significant security concern! There is a risk that perpetrators could replace the posted QR code with their own, which would redirect drivers to a fake website where they would unknowingly enter their credit card information. This could lead to unauthorized charges and potential identity theft. It is crucial for the shopping center to address this security concern by taking appropriate measures to protect drivers' personal and financial information. They must ensure that the QR codes posted are authentic and that there are no vulnerabilities that could allow perpetrators to manipulate the system. As an alternative solution, pay stations could be installed for shoppers to use instead of relying on their own phones.

33

12 Comments

Transform complexity into simplicity with Codemod's innovative migration solutions 🚀. Discover how to gain a competitive edge with seamless, efficient code migration at the Dev Innovation Summit 💡. Don’t miss this opportunity to redefine your approach! 💻

7

Happy Zivid SDK 2.12 release day 🥳 Dive into the improvements and new features: https://lnnk.in/hGj9 Excited to share the latest updates in SDK 2.12! Here's what's in store: 1️⃣ New Omni Engine v.2: Enhancing working distance and elevating point cloud quality. 2️⃣ Zivid on Nvidia Jetson Orin modules: Expanding possibilities with cutting-edge technology. 3️⃣ New ROI Feature: Dramatically reduces capture time by up to 400%. 4️⃣ Upsample/Downsample: Balancing capture speed and resolution perfectly. 5️⃣ Zivid Studio Enhancements: Introducing a suite of new features and interface improvements. 6️⃣ Camera Settings: Transitioning from experimental to fully integrated. 7️⃣ Important Announcement: Zivid One+ no longer supported by this and future SDKs. 🔍 Ready to explore the future of 3D vision technology? #SDK #3dvision #Zivid

5

This is a question many Data Engineers asked me: How are systems like Apache Druid, Apache Pinot, Clickhouse, etc... different from RisingWave? Aren't they all "real-time" and can't they do "streaming"? There are indeed fundamental differences between these systems and RisingWave. For OLAP Databases, their computation model is still executing in batch. They are "real-time" in a sense that their adhoc queries can execute really quickly. And they can do "streaming" in terms of ingestion, they can quickly stream in records, and run batch computation on them. For RisingWave, it's differentiated, because it does incremental rather than batch computation instead. So we can achieve lower latencies for streaming queries, and spend less compute to get the same results. That being said, streaming systems do not entirely replace OLAP/OLTP systems as well, there are many use cases where both can complement each nicely. Catch the talk to understand more about the fundamental differences, and how these systems can work cohesively. #risingwave #olap #streaming #database #realtime

17

Harness has teamed up with Google Cloud to integrate with the Gemini Code Assist tools. This innovative AI-driven tool is set to revolutionize the way developers work, offering smarter code generation and more advanced debugging capabilities. Brad Rydzewski, Vice President of Engineering, said, "This collaboration deepens our shared commitment to improving developer productivity — a core value at Harness — by delivering innovative AI-native solutions to accelerate and enhance software delivery." For more information, keep an eye out for updates: https://lnkd.in/ddBB4x3B #googlecloud #googlecloudpartners #GoogleAIpartner

3

When the Lights Go Out, the Learning Continues As the Puget Sound area faces power outages, I’ve taken the downtime as an opportunity to reflect and document my journey of learning system architecture. It’s a reminder of the importance of resilience—whether in systems or life. Here’s an article documenting key lessons and strategies I’ve gathered along the way. Welcome to any feedback 😀 #software engineer

12

1 Comment

I’ve officially launched my mentorship page on topmate.io Why Topmate over Calendly? It’s simple: Topmate isn’t just about scheduling - it’s a complete ecosystem for meaningful conversations. The platform is tailored for deeper engagements, seamless integrations, and creating real value for those seeking guidance. Big thanks to Pritam for the stellar recommendation. I’m kicking off the QueKey Integration and Technical Assistance initiative. This is a hands-on effort to help businesses streamline processes and unlock their full potential using QueKey. Whether it’s technical hurdles or strategic optimization, this program is here to deliver real, actionable solutions. Topmate is the perfect platform for these efforts because it’s designed for efficiency, scalability, and delivering results. It’s not just a tool - it’s a partner in creating impact. Connect me here: https://lnkd.in/g5WqjTbi Let’s innovate, solve, and grow together. #Mentorship #Innovation #QueKey #Topmate

24

2 Comments

LLMs and LangChain: The Dynamic Duo 🦸‍♀️🤖 Just finished a fun project where we built a Q&A bot using these two powerhouses. It's like having a personal AI assistant that's always ready to answer your burning questions! 🔥 Here's what I learned: LLMs are wizards: They can conjure up answers from thin air (or, uh, the internet). 🧙‍♂️ LangChain is the glue: It holds everything together and makes it easy to use. 🔗 Endless possibilities: This tech can be used for anything from customer support to writing catchy social media posts. 📝 Want to see it live (obvio a test server link 😛 ): https://lnkd.in/dZNAVCSV Note: it's a render link, they will take down the instance because of no use, you can hit the link and wait for two minutes, they will spin up the instance and you can try it out. Want to see the code? Check it out here: https://lnkd.in/dzjh4r47 #AI #MachineLearning #NLP #LangChain #LLM #Python #Developer #FunProject"

22

What an incredible experience at #KubeCon North America 2024! 🌟 I had an amazing time at the #OpenSearch booth connecting with community users and contributors. The energy and excitement around OpenSearch moving to the Linux Foundation was truly inspiring! 🚀 It was fascinating to hear how OpenSearch is being used in so many ways—from self-managed clusters and managed offerings to partner solutions. The versatility of OpenSearch shone through in conversations about use cases like hybrid searches, #RAG pipelines, #Observability, #SIEM, and even as a #tracing solution. A big highlight was the feedback on OpenSearch Dashboards v2.18—our latest updates on workspaces, SQL - PPL support in discover are clearly resonating with users, and it's always gratifying to see the impact of our work. Lastly, the OpenSearch roadmap session was a proud moment! Reflecting on the milestones achieved this year and the exciting vision for 2025 in the OpenSearch Suite reinforced just how far we’ve come as a community. Thank you to everyone who stopped by, shared their stories, and contributed to making OpenSearch better every day. Here’s to the future of open source search and analytics! #KubeCon2024 #OpenSearch #LinuxFoundation #OpenSource #SearchAnalytics #Community #Observability #OTel

86

1 Comment

Recent events, such as the tragic murder of a CEO over unpaid dues for three months, highlight a critical gap in our system. This could have been prevented if proper labor courts were established to address the grievances of the salaried class in a timely manner. Unfortunately, we live in a banana republic where, when it comes to applying taxes, the salaried class is the very first target. However, when it comes to providing legal support, there is no proper public defender system in place. Either you hire a lawyer and endure the daunting choice of lengthy court battles taking years, spending more than the owed dues in lawyer's fees and other expenses, or you take the other extreme path and stab the CEO. There is nothing in between.

11

2 Comments

AWS re:Invent was packed with GenAI workshops and sessions this year. As I had hoped and expected, AWS killed it with anything and everything related to storage and compute (see dsql, knowledge bases, Graviton training, …), but AWS, like many, is still only barely scratching the surface on business automation with AI. To summarize, AWS now supports AI Agents on Bedrock. Imagine a supervisor LLM who, when tasked with a complex assignment, breaks it down into small steps and spins up Lambdas of specialized agents to complete the task. This sounds great until you notice that all the demos were scoped to do high-level analytics or create drafts for creative marketing campaigns. This bothers me for multiple reasons: 1. A Lambda for every action is reminiscent of the microservice bloat era—not everything needs to be a fully managed microservice. 2. LLMs were supposed to unlock human creativity, not replace it with AI slop. 3. LLM-only agents cannot be trusted to do write operations! Talking to one of the big bank partners at the event, I learned that businesses still don’t trust LLMs with automation. I had noticed the same problem of trust around a year ago when I was trying to use LLMs to automate DevOps. No engineer was willing to pay to give up control of the critical operations to LLMs, and that is what led me to Kognitos! Kognitos had developed a pre-LLM natural language interpreter, therefore fundamentally eliminating the risk of hallucinations. Over the past year, we have been working on making the platform more flexible and easier to use (see demo in my profile). Improvements are not motivated by some short-term goals set by our partners but by our long-term vision to enable full lifecycle automation without the use of conventional programming. We are looking for new testers (especially those with automation backgrounds, e.g., RPA) and companies looking to test out the latest releases. PM me if interested.

24

1 Comment

A Day at the Apache Kafka 🚀 Today, I had the pleasure of attending an in-person Apache Kafka hosted by CRED in Bangalore. Here’s my feedback on this incredible event: 📍 Venue: Cred Aero, Bengaluru 🔹 Agenda Highlights: - @Vikas Sharma , CRED : Talk: Running Kafka Optimally in Production: Practical Usage Feedback: Vikas shared practical insights on optimizing Kafka in production environments, with valuable strategies for resource management. - Udayaram Kammara , Bazaarvoice : Talk: How Bazaarvoice Utilizes Kafka for Both Stream and Batch Processing Scenarios Feedback: Udayaram's session on Kafka's use at Bazaarvoice was enlightening, covering real-world use cases and challenges. - Diptiman Raichaudhuri , Confluent : Talk:Stream Processing for the Non-JVM World: pyflink and kafka-python Feedback: Diptiman's explanation of Apache Flink's architecture and the demonstration of kafka-python and pyflink were fascinating and beneficial for non-JVM users. What I Learned: - Effective strategies for running Kafka in production. - Real-world use cases and challenges in stream and batch processing. - Leveraging pyflink and kafka-python for stream processing. Overall, this meetup was a treasure trove of knowledge and inspiration. The sessions were well-organized, and the speakers were exceptionally knowledgeable. I’m excited to implement these learnings in my projects and continue exploring the potential of Apache Kafka and related technologies. 🙌 #ApacheKafka #KafkaMeetup #CRED #Bangalore #EventStreaming #StreamProcessing #TechCommunity #Networking #Python #Flink #DataPlatforms

24

2 Comments

NVIDIA's CEO, Jensen Huang, recently praised Elon Musk, calling him "superhuman" for his incredible work on a powerful new computer at his company, xAI. Musk’s team built this supercomputer, named "Colossus," in just 19 days — a project that usually takes up to four years to complete. Huang admired Musk's strong engineering skills and his talent for bringing the right people and resources together to get things done quickly. This computer, Colossus, will help xAI dive deeper into advanced AI projects, making it possible to work with complex models and learn even more from data. The project is a big step forward for xAI and highlights the strong partnership between Musk’s team and NVIDIA, showing just how much Musk’s fast-paced style is respected by leaders in technology. #ElonMusk #AI #NVIDIA #xAI

7

8 Comments

Longhorn Electric Racing Is Building the Future of Electric Vehicles with Cadence Tools. Longhorn Electric Racing is a student-run racing team at the University of Texas, Austin. Not only are they building the best race cars, but they’re also building the future generation of engineers. Their electric vehicle team is one of their most recent teams, founded in 2016. Some of the challenges they face in designing an electric car are making sure it can go as fast as possible, making sure it has sufficient cooling, and making models effectively. With the help of Cadence CFD and Cadence OnCloud, they’re reaching their goals much more efficiently. https://ow.ly/phqj50RuqK4

19

I just finished building my latest AI app and entry into the NVIDIA + LangChain Generative AI Agents Developer Contest (https://lnkd.in/gYtKqB9C)! Check out the demo video below, or continue reading for a deeper technical dive. CLARA 2 is a 3D-rendered AI agent that will present your PowerPoint for you. It's powered by multiple machine learning models: machine vision with OCR/object recognition, a very powerful open-sourced Mixtral 8x22b, TTS (text-to-speech), and real-time audio-->animation streaming powered by Audio2Face. This project marks a major evolution from the original CLARA app I created a few months back. Previously, with CLARA "1", I handled all of the ML computations locally, but with CLARA 2, I needed to optimize system resource usage (namely, VRAM) by utilizing the NVIDIA NIM API with the LangChain framework for calls to Mixtral 8x22b for inference. The NVIDIA NIM + Langchain abstraction SIGNIFICANTLY reduced the complexity of the infrastructure setup that would have been needed for CLARA 2 otherwise. Not only that, but building CLARA 2 simply would not have worked running totally locally, as smaller LLMs cannot make sense of the large context of an entire PowerPoint presentation that CLARA 2 needs to work. Worse, larger LLMs are impossible to run in a performant fashion on consumer hardware due to VRAM limitations. As a developer, if you can't get access to the hardware you need to develop locally, you want a service that doesn't take a lot of time to set up, consistently works, and is fast during inference. Frankly, NVIDIA NIM + LangChain checked all those boxes. During my development of CLARA 2, I made over 500 calls to Mixtral 8x22b, and every call was performant and worked as expected. Kudos to NVIDIA and LangChain for the dev-friendly experience using the NIM API + the LangChain framework integration. Now, that's not to say everything in CLARA 2 ran on APIs. All the CLARA 2 logic is local, and, due to NVIDIA's backend doing the heavy-lifting of Mixtral inference, my 4090 was freed up to run the other VRAM-heavy components such as the inference performed by Audio2Face and the real-time rendering of a 3D MetaHuman in Unreal Engine. Anyway, if you made it this far, thanks for reading. If you want to learn even more, check out the CLARA 2 GitHub page: https://lnkd.in/gPxYqyvg. There I've provided detailed documentation so that you can set up CLARA 2 yourself. Feel free to use what I've built as inspiration for your own AI Agent, and reach out if you have any questions. Cheers. #NVIDIADevContest #LangChain

101

3 Comments

With Google's recent release of their quantum chip, Willow, on December 9, 2024, I am reminded of an anecdote I'd like to share. I remember seeing a pretty sunrise in 2021, sitting by the beautiful Sukhna Lake(Chandigarh), watching the sunrise with Prankush Garg, a sophomore in Applied Mathematics at IIT Roorkee. The air was filled with excitement as we delved into discussions about quantum computing. IBM had just rolled out its Quantum Cloud program, giving enthusiasts like us a chance to experiment with real quantum computers. For two self-proclaimed physics aficionados, it was a dream come true to transition from bits to qubits, running algorithms like Shor’s on these rudimentary but groundbreaking machines. Of course, we quickly realized that quantum computing was still in its infancy. Practical, real-world applications felt far away. Yet, we couldn’t help but dream. “What if?” we asked ourselves, imagining how quantum computing could revolutionize modern infrastructure—finance, technology, and healthcare—unlocking the power to brute-force solve problems that today’s computers couldn’t even touch. NP problems? No problem (someday). Fast forward to December 2024, and I woke up to goosebumps reading Google’s announcement of "Willow"—their cutting-edge quantum processor. Google claims that Willow performed a benchmark computation in under five minutes that would take the world’s fastest supercomputer 10 septillion years. Yes, you read that right—10²⁵ years—a number that dwarfs the age of the Universe itself. While Google reassures us that today’s cryptographic systems aren’t quite at risk yet (with a playful wink in their statement), it’s clear we’re standing on the brink of something transformative. Living in an era that’s witnessing two century-defining revolutions—AI and quantum computing—feels nothing short of a blessing. The future isn’t just knocking; it’s here. Here’s to the what-ifs turning into what-is. 🚀✨ #QuantumComputing #Innovation #AI #Technology #ShapingTheFuture

18