“I had the pleasure of working with Dan on several projects, including the development of the Python SDK at Edge Impulse. Daniel's strong foundation in machine learning and his ability to provide clear customer stories and detailed code reviews were instrumental in creating a robust and user-friendly interface. His passion for edge AI and talent for simplifying complex concepts greatly benefited our team. Daniel's leadership of the ML team at Edge Impulse resulted in innovative models and systems that empower users to build edge AI products addressing real-life problems. His ability to articulate the technical vision and motivate the team was key to our success, fostering a collaborative and growth-oriented environment. I am grateful for the opportunity to have worked with Daniel and highly recommend him for his expertise, leadership, and visionary thinking in the field of machine learning and edge AI.”
Daniel Situnayake
Mount Pleasant, South Carolina, United States
3K followers
500+ connections
About
Daniel Situnayake is Director of Machine Learning at Edge Impulse, and a technologist…
Activity
-
NVIDIA released the new Jetson Orin Nano Super today, and we were fortunate to receive one in the mail! 🚀 Here's a quick unboxing video from…
NVIDIA released the new Jetson Orin Nano Super today, and we were fortunate to receive one in the mail! 🚀 Here's a quick unboxing video from…
Liked by Daniel Situnayake
-
Our team will be at #CES2025 and is looking forward to meeting with anyone interested in building edge AI applications! There are demos featuring…
Our team will be at #CES2025 and is looking forward to meeting with anyone interested in building edge AI applications! There are demos featuring…
Shared by Daniel Situnayake
-
I was lucky enough to catch up with Ann Spencer to talk about why speech to intent is going to be so important, here's our conversation.
I was lucky enough to catch up with Ann Spencer to talk about why speech to intent is going to be so important, here's our conversation.
Liked by Daniel Situnayake
Experience
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Education
-
Birmingham City University
-
Received the TIC Undergraduate Computer Networking Prize as the top performing Networking student of 2007.
-
-
Volunteer Experience
-
City Contact for San Francisco Bay Area
Labour Party
- 2 years
Politics
The Labour Party is the UK's largest political party. I run meetings and events for members resident in the San Francisco Bay Area.
Publications
-
AI at the Edge: Solving Real World Problems with Embedded Machine Learning
O'Reilly
Edge artificial intelligence is transforming the way computers interact with the real world, allowing internet of things (IoT) devices to make decisions using the 99% of sensor data that was previously discarded due to cost, bandwidth, or power limitations. With techniques like embedded machine learning, developers can capture human intuition and deploy it to any target--from ultra-low power microcontrollers to flexible embedded Linux devices--for applications that reduce latency, protect…
Edge artificial intelligence is transforming the way computers interact with the real world, allowing internet of things (IoT) devices to make decisions using the 99% of sensor data that was previously discarded due to cost, bandwidth, or power limitations. With techniques like embedded machine learning, developers can capture human intuition and deploy it to any target--from ultra-low power microcontrollers to flexible embedded Linux devices--for applications that reduce latency, protect privacy, and work without a network connection, greatly expanding the capabilities of the IoT.
This practical guide gives engineering professionals and product managers an end-to-end framework for solving real-world industrial, commercial, and scientific problems with edge AI. You'll explore every stage of the process, from data collection to model optimization to tuning and testing, as you learn how to design and support edge AI and embedded ML products. Edge AI is destined to become a standard tool for systems engineers. This high-level roadmap will help you get started.Other authors -
-
TinyML: Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers
O'Reilly
Deep learning networks are getting smaller. Much smaller. The Google Assistant team can detect words with a model just 14 kilobytes in size--small enough to run on a microcontroller. With this practical book you'll enter the field of TinyML, where deep learning and embedded systems combine to make astounding things possible with tiny devices. As of early 2022, the supplemental code files are available at https://oreil.ly/XuIQ4.
Pete Warden and Daniel Situnayake explain how you can train…Deep learning networks are getting smaller. Much smaller. The Google Assistant team can detect words with a model just 14 kilobytes in size--small enough to run on a microcontroller. With this practical book you'll enter the field of TinyML, where deep learning and embedded systems combine to make astounding things possible with tiny devices. As of early 2022, the supplemental code files are available at https://oreil.ly/XuIQ4.
Pete Warden and Daniel Situnayake explain how you can train models small enough to fit into any environment. Ideal for software and hardware developers who want to build embedded systems using machine learning, this guide walks you through creating a series of TinyML projects, step-by-step. No machine learning or microcontroller experience is necessary.Other authors -
Languages
-
English
Native or bilingual proficiency
Recommendations received
1 person has recommended Daniel
Join now to viewMore activity by Daniel
-
This is contextual AI running 100% on-device — no cloud, no offloading. Just a private, responsive, voice assistant 🤯 Imagine any appliance — in…
This is contextual AI running 100% on-device — no cloud, no offloading. Just a private, responsive, voice assistant 🤯 Imagine any appliance — in…
Liked by Daniel Situnayake
-
Congratulations to Reka Meszaros and Hanna Selim at Aalborg University on their healthcare AI project using Edge Impulse, recently demonstrated to…
Congratulations to Reka Meszaros and Hanna Selim at Aalborg University on their healthcare AI project using Edge Impulse, recently demonstrated to…
Shared by Daniel Situnayake
-
I'm looking forward to being a keynote speaker at the Generative AI for Edge Computing Symposium, at the AAAI 2025 Spring Symposium from March 31 to…
I'm looking forward to being a keynote speaker at the Generative AI for Edge Computing Symposium, at the AAAI 2025 Spring Symposium from March 31 to…
Shared by Daniel Situnayake
Explore more posts
-
Ziaul Kamal
Do Open Source Obligations Change with Packaging Ecosystems? https://lnkd.in/gfTxqiYr The world of open source software has evolved significantly over the years. What started as a way for developers to freely share code has turned into a complex ecosystem with its own set of norms, expectations and even obligations. Much has been written about what companies should and shouldn’t expect or demand from maintainers of open source projects, and related compensation considerations. To be clear up front, I fully agree with the general consensus that open source code is freely given and that maintainers who simply publish their code have no obligations beyond what the license states. But I want to explore a controversial but important topic: what we as a community can and should expect from open source projects and maintainers who participate in packaging ecosystems. When a maintainer takes the extra step of participating in a packaging ecosystem, I believe that triggers a new realm of obligations and social contracts — and that more public discourse and debate around this topic would be especially healthy for the open source industry at this moment in time when the XZ Utils and similar exploits are creating a special urgency around trust and transparency in build systems. This is a conversation that is way overdue, and at the crux of the modern software supply chain security movement. New Distribution Models, New Responsibilities Let’s begin by taking a quick look at how open source has evolved. Historically, the responsibilities of building and distributing open source software were typically separated. Developers would publish their raw source code, then other communities like Linux distributions would take on the task of packaging and distributing that code to end users. These distribution communities, while often also open source, operated under their own policies and social contracts with users. However, the rise of language-specific package managers like npm, PyPI and RubyGems blurred this line between “code vs. packages.” Suddenly, publishing packages became accessible to any developer. While this greatly accelerated open source development, it also introduced vastly different security guarantees compared to traditional Linux distribution packages. Many developers, myself included, didn’t fully understand these trade-offs at first. These package managers made publishing accessible to anyone. They brought similar UX, but vastly different security guarantees. My first memories of Python involved struggling to get NumPy working and not understanding the difference between apt-get install and pip install, except that one worked and the other didn’t. I didn’t realize or understand the security trade-offs that were involved. The package managers look and feel the same. npm install feels a lot like AppGet install or PIP install, for example but they do completely different things from a trust perspective. While Linux distribution pac...
-
Ali Ahmad
💡 Top No-Code Agencies in 2024—Innovation at Your Fingertips! 💡 Say goodbye to traditional development complexities! From startups to enterprises, these top no-code agencies are making it easier than ever to launch MVPs and full-scale applications. Check out our latest blog for a deep dive into the best agencies of 2024, and see how they’re reshaping the landscape of app development! 📈✨ 👇 Discover which agency is right for your next big project! Want to see how no-code can transform your project? Reach out for a free consultation with NoCodeProvider! #nocoderevolution #techinnovation #appdevelopment #digitalsolutions #businessgrowth #nocodeprovider
-
Josiah Parry
🔍 A new R vulnerability was announced recently 🐞 The vulnerability is easy to avoid: upgrade R and don't load random files from the internet. It doesn't scare me because R is a real programming language and there are infinite things you can do with it. You can hide just about anything in an R package using things like: - .onLoad() to run arbitrary code when a package is loaded - include pre-compiled code - use side effects in functions - hide malicious functions in S3 generics https://lnkd.in/g5nUvjcY #rstats #security
6 Comments -
Ziaul Kamal
A React-Based Open Source Tool for Creating Data Tables https://lnkd.in/gfTxqiYr AG Grid‘s developer relations lead, James Swinton-Bland, doesn’t see TanStack Table as a competitor. TanStack Table is an open source library for creating tables in the React framework, formerly known as React Table. On the contrary, the data grid company sponsors TanStack Table and considers TanStack a partner. “Our founder, Niall [Crosby], who sadly passed away recently, was a big fan of supporting open source communities and open source projects, and a big fan of supporting Tanner [Linsley] specifically,” Stinton-Bland said. “Tanner’s React Table solution is headless. Essentially that means that you have to build out the UI yourself, and you have to style everything, and you have to do a lot of legwork to get things working.” AG Grid, an open source solution, is a feature-rich JavaScript data grid library used to build interactive and sophisticated data tables or grids in web applications. A JavaScript grid is a UI component used to display and manage tabular data in a web application. It provides a structured way to present information in rows and columns, making it easy for users to understand and interact with the data. AG Grid provides a way to put these grids into your web applications, he added. It’s essentially like React Table but with “batteries included,” he added. “The simplest way I can describe that is it’s built to be basically Excel for web developers,” he said. “Essentially, it lets you put a very powerful table into your web applications.” Last year, the company launched AG Charts, a charting library created as a way to power the integrated charting features of AG Grid, where it’s possible to highlight data and build charts from it. “In order to power that functionality, we essentially had to build a whole charting library to do that, and we took the decision last year to turn it into its own product, essentially,” he said. “Now AG Charts is a completely standalone JavaScript charting library with support from React, Angular and Vue, much the same as AG Grid.” The two products are built to work seamlessly together, but they can also be used as standalone applications, he added. Support for Frameworks For both products, the core business logic is written in JavaScript, which allows the solutions to be very fast, Stinton-Bland said. AG Grid and AG Charts support React, Angular, Vue and JavaScript with full TypeScript support. Solid was supported, but that has been deprioritized due to less interest than anticipated, he said. However, Solid does offer its own wrapper for AG Grid, he added. “We’re not restricted by any of the restrictions of the individual frameworks, but then we actually provide full support for those frameworks on top of that, and it’s not a thin wrapper layer or anything like that,” he said. “All of the rendering is done in pure React, for example. In that way, you can get all of the benefits of performing...
-
Prof Bill Buchanan OBE FRSE
That pesky ECDSA signature refuses to go away, and we are struggling with it in a distributed world of trust. Luckily EdDSA is much more scalable, and now used in many distributed signature applications (and where we can split the private key up into secret shares and distribute them over several nodes). But we have a problem ... elliptic curve methods are going to be cracked by quantum computers. So can we take our PQC signature methods and scale them to distributed signatures? Well, no, as Dilithium doesn't support it. But Raccoon can, and also has better side channel resistance (and with a similar performance and key size to Dilithium. So, meet Raccoon: https://lnkd.in/d2QMHSM9
2 Comments -
Mason Sharp
Included in this article is the mention of vector databases. Unsurprisingly, Michael Stonebraker and Andy Pavlo feel there is not a need for vector-specific databases. They believe relational databases will adopt their features, which is already happening. Check out the details here: https://lnkd.in/gyq_4WjP
1 Comment -
Sir Angelo Reale, KPO
Consolidated my Notion homepage on a custom domain. I'm surprised it's not yet a native feature. I've realised that while pointing my personal website (https://angeloreale.com), I had lost most of my few years old SEO authority. So then, ... we just build it again. :) Recommending this Notion wrapper, and here is my referral link (I get some bucks if you sign-up using it): https://lnkd.in/dNKCuppi
-
Tristan Lombard
Shoutout to our Claude fans - Anthropic just released a GitHub repo with 5 free courses (including prompt engineering techniques): 👩💻 Anthropic API fundamentals - teaches the essentials of working with the Claude SDK: getting an API key, working with model parameters, writing multimodal prompts, streaming responses, etc. 👩💻 Prompt engineering interactive tutorial - a comprehensive step-by-step guide to key prompting techniques. [AWS Workshop version] 👩💻 Real world prompting - learn how to incorporate prompting techniques into complex, real world prompts. [Google Vertex version] 👩💻 Prompt evaluations - learn how to write production prompt evaluations to measure the quality of your prompts. 👩💻 Tool use - teaches everything you need to know to implement tool use successfully in your workflows with Claude. Free continuously learning starts here: https://lnkd.in/gFGQWG4P #PromptEngineering #Dev #SoftwareDevelopment #API #SDK #AlwaysBeLearning
3 Comments -
Leo Epstein
GenAI tools like GitHub Copilot are here, and are making a major difference (when used properly) Looking forward to this conversation with Laura Tacho and Nathen Harvey discussing some of the ways not to leverage GenAI tooling, as well as how DORA is thinking about GenAI in 2024. https://lnkd.in/gAGeRixh
-
Errol Schmidt
Enums enhance code readability and clarity. They define named values for model attributes, making them ideal for representing statuses, stages, and states. An article from the team at reinteractive explores the powerful features of enums, particularly how they make code readable and self-explanatory … https://buff.ly/4cKGWiu #RubyOnRails #Programming #Enums #reinteractive
-
Rohit Khare
Summer Camp for Cybersecurity week made a lot of waves, with a number of #cloudsecurity vulnerabilities debuting in Las Vegas around failures of #AuthZ logic and delegation from human to non-human identities. Please subscribe to our volunteers’ latest issue of our free weekly newsletter of all things #authorization!
-
Dexter Horthy
fun post from Harrison Chase on cognitive architectures (link in comments) - does a great job of articulating the difference between human-defined workflow DAG vs. letting the LLM drive the workflow In a human-defined DAG you get greater control and determinism, but I've talked to folks who have experimented with agent-defined DAGs, and many of them remark things like "the agent found a solution that we never would have come up with" - I think as LLMs get smarter, there's a ton of upside in taking the rails off the workflow which one do you prefer? #ai #llms #agents
6 Comments -
Jeremiah Small
Free software is free as in libre, not free as in gratis. Free software advocate and GNU founder Richard Stallman has quipped, "Think free as in free speech, not free beer." In the world of free open-source software (FOSS) many large popular open-source projects are well-funded professional organizations. For example, React is a FOSS project maintained by Meta. But countless FOSS projects rely on small teams of volunteers to keep them vibrant and secure. At Soliant Consulting we like to financially support third-party small-but-essential projects that form the backbone of our work. One such project is MikroORM – a TypeScript ORM that has become an integral part of our development stack. MikroORM is maintained primarily by the creator, Martin Adámek. MikroORM has a design philosophy similar to those of Doctrine and Hibernate. The software recipe we reach for when building closed-source custom cloud native applications is built on AWS (Amazon Web Services) and relies on many FOSS projects. These range from massive projects like Node itself, to small projects like MikroORM. Cloud infrastructure allows us to automate everything from development bootstrapping through continuous delivery to production, and FOSS gives us the freedom and flexibility to focus on our customer’s requirements. If you are building closed-source projects with FOSS, consider supporting the projects you rely on. It is a small gesture that keeps innovation alive. Remember you are free to do whatever you like with FOSS, but it is not free as in beer. Check out MikroORM and consider sponsoring them if you find value in their work: https://lnkd.in/gBiSzca4 #opensource #foss #softwaredevelopment #cloudnative #typescript #AWS #MikroORM
-
Tihomir Opacic ✨
I just tried ChatGPT's new Canvas feature, and I think it will be useful. Canvas is a practical interface that lets you work directly on code or any textual document without the limitations of a traditional LLM chat window. It is similar to Anthropic's Claude Artifacts, but it seems to be an improved version of Artifacts. Here's what I find most useful out of the box: - Coding shortcuts: You get Review code, Add logs, Add comments, Fix bugs, and Port to a language at the click of a button. - Versioning - Chat inline with a piece of code - I can make edits directly in the Canvas tool I’m curious to see how it evolves! https://lnkd.in/ewQBBYZw #AI #ChatGPT #Canvas #Coding #Productivity
1 Comment -
Carrie Morales
Google's IO 2024 was last week with announcements about Gemini and other AI developments. The two most exciting improvements will be Talkback using Gemini for image descriptions and Gemini having the ability to process video. This could have SO many implications for me and my fellow #blind and #VisuallyImpaired people out there. I share more highlights from Google IO an #accessibility perspective in my latest video! ⬇️ #LowVision #Disability #A11y #GoogleIO #AIforGood
-
Julia Renouard
Feedback loops - cognitive load - flow state. This is my favorite part of how to think about developer productivity. Laura Tacho has provided a fabulous guide on how to think about up-leveling your team. It is about the developer experience but also with quantifiable measures. This guide provides more color on the multiple dimensions you need to consider for making your teams more effective and realizing that there is no one-size-fits-all measure or solution.
1 Comment -
Dror Gill
Anthropic just launched Claude 3.5 Sonnet, which beats both OpenAI GPT-4o and Anthropic’s previous paid model Claude 3 Opus on most benchmarks. Claude 3.5 Sonnet can execute and debug code, and has improved image understanding capabilities, including text transcription. A new feature called “Artifacts” enables users to view generated code, documents and websites in a separate window next to the chat. Claude 3.5 Sonnet is available for free to all users, but paid users get higher usage caps. Did Anthropic just take the lead back from OpenAI? It seems that ChatGPT’s Code Interpreter tool, which used to be one its main advantages, has finally met its match, but ChatGPT still has the advantages of real-time Internet access and custom GPTs which Claude lacks… What do you think? https://lnkd.in/d_hk86cf
3 Comments -
Richard Christopher
I have spent half of my life thinking about governance, political and economic systems, and sociology. I've often played mental chess in my head dissecting the pro's and con's of various types of governance - comparing the political theory on paper to different sorts of execution in practice, and looking at how these systems support the masses as well as the individuals. It's 2024 and we haven't figured out voting, asset management, education, finance, healthcare, global diplomacy, or even freedom of information. We've been perverted by generations of oligarchal tyrants and totalitarianistic monopolies - while being led to think that our only options are Communism, Socialism, Democracy or Fascism. I took the time, after many years of picking apart political systems, to build an outline of how we could have a system that has all the benefits of communism (yes, there are benefits - on paper), with the complete freedom of anarchy.. all of the benefits of capitalism, with the protections against oligarchies and extreme socialist republics. While it looks good on paper, it would take a complete mind-shift of the masses in order to incorporate something like this, and would first take us being honest with ourselves, and being willing to admit that we fucked up. An addict cannot recover unless he first admits that he has a problem, and we have all become addicted to convenience, and short term dopamine hits.. and we are surely going to pay for it if we don't make some changes. Without any further delay, here is an outlined proposal of what a new type of governing system could look like for the people of the world: Libertocracy. The governments and corporations of the world today only have authority because we grant them that authority. You cannot build a house ontop of a broken foundation. One must be willing to build a new system to make the old one obsolete. It's time for a change. #government #theory #politicaltheory #politicalscience #systemdesign #politics #globalthreat #tyranny #totalitarianism #communism #socialism #fascism #democracy #capitalism #marxism #libertocracy #freedom https://lnkd.in/eCspHbPP
-
Keith Brings
#elixir #ecto #seeds Here is a little utility standardizing some logic i've used for a while to manage incremental seed updates for ecto in elixir. It can track applied seed sections, add condition blocks that only run in specified environments, and blocks that are queued until a referenced seed or list of seeds have been applied. https://lnkd.in/g2pSmQE2
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More