G Wu

Collision Conf is one of the biggest tech+AI events in North America and I am looking forward to learning more in this exploding space. If anyone is heading that way, feel free to ping and we can connect! #AI #generativeAI #digitaltransformation

4

Some interesting thoughts in the inexorable march forward of each AI generation, but also the limits to where AI really shines. In terms of repurposing exisitng content, there seem to be some real value adds here - but exactly how useful this 'professional' podcast really is remains to be seen.

10

2 Comments

Cohere CEO Aidan Gomez says that #1 blocker for Enterprise GenAI use cases is data privacy and #2 blocker is the cost (implies ROI). He also explained what Canadian AI startup, Cohere does and how they make money. A must watch for anyone who is interested in Enterprise GenAI use cases. He also said that GenAI fear mongering is cheap ways to get attention (I am paraphrasing, he didn’t use the word cheap). Quite a bit commentary on Google Cloud AI developments as well. Other mentions Oracle and Massachusetts Institute of Technology. Aidan comes across as a mature tech leader for his young age. He sounds more realistic and less me too, in GenAI hyped era. Aidan was talking to CNBC’s Steve Kovach. https://lnkd.in/g36SFYfn ____________________ Last month, Cohere has raised $450 million in funding from returning investors such as Nvidia and Salesforce Ventures, as well as new investors including Cisco and Canadian pension fund PSP Investments cc David Vellante David Linthicum John Furrier Ron Miller Shelly DeMotte Kramer Robert Hof

15

I'm excited that this week at the CAIAC - Canadian Artificial Intelligence Association conference George Wang will be presenting a paper Adaptive Learning Rates for Gradient Boosting Trees, based on work he did as an intern at Kinaxis last fall, in collaboration with Christopher Wang, Yunfei O., Behrouz Soleimani, PhD. There are 3 reasons this is both impressive and cool: 1. They combine heuristics, machine learning, and optimization to speed up the training for an important algorithm, Gradient Boosted Machines, widely used in both academia and industry. Faster training improves model performance. 2. This paper is just one of the many sparks of innovation in #artificialintelligence coming out of Kinaxis, where > 50% of our patents filed in the last several years have been in this space, because we continue to push the boundaries by inventing new solutions like this one to solve problems our customers face. We've been delivering AI to the market since 2018 and were the first to provide solutions powered by AI to address both demand and supply. We are rooted in Canada, which brought the world out of the last AI winter in 2012 when a team from the University of Toronto won the famous Imagenet competition by blowing the lid off previous results. Their submission was layers of neural networks called deep learning, which later became the foundation for the #generativeAI revolution we are seeing today. 3. We have an outstanding intern program in #machinelearning at Kinaxis, where interns work on real problems, guided by senior staff. As a testament to letting the students shine, when their paper was accepted they decided to have George present it. So check out the paper if you're at the conference, but if you miss the event you can't miss the fact that we are actively speaking and sparking on AI. In fact, just last week Chantal Bisson-Krol spoke in Edmonton at the #upperbound conference on AI-Powered Strategies for Resilient Supply Chains and I spoke at the Reuters Supply Chain USA event on Embedding AI in supply chain to deliver end-to-end value.

43

2 Comments

Important fact about most startup (and VC) success stories. It is convenient to overlook this, but the best companies and products are not built on any specific timeline. The best people win, and that is the only thing that has consistently and repeatably truly won over the test of time.

39

AI memory is fundamental for conversational AI and AI agents. It is critical for building personalized experiences and improving reasoning. Check out Letta’s course to understand how to use AI memory to enhance the customization and quality of your AI app.

15

With the launch of Llama 3.1 405B model, it's finally feasible to deploy a fully private full-stack GPT4-class Gen AI app, a game changer for regulated industries. For many use cases, the 70B parameter model may be good enough, and it can be deployed on a single 8-GPU server, well within the reach of enterprise IT. Exciting times! Obviously, all Gen AI apps need orchestration ... https://lnkd.in/ef3GR6sS

12

6 Comments

The introduction of KANs is super exciting, showing a big leap in machine learning for specific uses. Yet, this doesn't mean KANs outperform traditional LLMs in every aspect. For general language processing tasks such as text generation, comprehension, and translation, LLMs still lead the way due to their extensive training on diverse datasets. They also excel in scalability and deployment, having been widely tested and used in real-world applications. This versatility makes LLMs invaluable across various industries, providing powerful tools for many businesses and developers. Meanwhile, KANs shine in areas where precise function approximation is critical, particularly in scientific and engineering applications. They offer better interpretability, which is essential in scenarios where understanding the model's decision-making process is vital. Additionally, KANs are more adaptable to new tasks without losing previously learned information, effectively handling the issue of catastrophic forgetting that plagues traditional neural networks. It’s important to note that when I say KANs excel in "niche" fields, I don’t mean these areas are small. Rather, they are focused yet represent large potential markets with significant impacts and opportunities. In essence, while KANs bring new strengths, especially in niche fields, LLMs remain the go-to for broader language-related tasks and practical applications, illustrating the complementary roles of these technologies in the AI landscape.

8

Startup decks vs Business plans. It's accepted that a startup needs a deck. Dax Dasilva of Lightspeed Commerce made an important point on 20VC. Writing a business plan - including financial projections - is an important step. It requires discipline. It provides a rigorous framework for thinking deeply about the problem you're trying to solve. Harry Stebbings countered that he doesn't even want to see a deck since it's all dreams. We spent a lot of time and energy on our business plan with the expectation that we would show it to no one ever. The business plan was part of our roadmap for the company. We researched customers - including having 50+ conversations. We wanted to understand them: their background, their interests, their presence on social media. We also wanted to understand how they were evaluated by leadership and how integrated they were with the business of the company they were protecting. We researched pricing both with potential customers and other products in our industry to find out if OpEx or CapEx was a better fit. How did they buy today? How would they like to buy? We did market research to understand the verticals and map the right entry point - we were wrong but that doesn't reduce the value of the research. We looked at competitors in depth to understand what they were seeing that we weren't. Dax is right. This is an invaluable document that should be updated every year - the company is changing, the market is changing, there's more feedback about price and ICP. It's not for sharing with the outside world. It's a map for driving forward. Decks won't get you investment - VCs invest in people. We make decks but we know they don't tell our story. The deck is a calling card. Dax and Harry are both right. https://lnkd.in/e25M82xY

15

1 Comment

Weekly Update 8.2.24 ⚡ • One month ago, I posted about how we shifted our operating cadence to better manage the tension between dedicated deep work time vs. managing increased merchant onboarding volume as we scale our Shopify App. A few shifts we made as an organization: 1. Engineering cycles — 2 week sprint on new functionality, 1 week focused on maintenance/tech debt. This ensures we continue shipping at a face clip while dedicating 1 week towards housekeeping. So far so good. 2. Dedicated deep work time — Jennifer is leading the charge by carving out 9 am - 1pm on select days for sacred deep work time, allowing her to focus on her top priority. Others on our team will follow lead by carving out dedicated deep work time and silencing notifications during these hours. I'm delinquent on this! I'll follow suit on this next week. 3. Product-led engineering — we've been inspired by companies like PostHog that champion product-led engineering and people like Arnaud Lassignardie, who takes extreme ownership. Our engineering team at TryNow increasingly makes product decisions, interfaces directly with merchants, and pulls forward our product roadmap. Typically, Product Managers define the "What to build" and Engineers define the "How to build." At TryNow, we're increasingly relying on Engineering to define select components of the "What" in addition to the "How." This level of ownership makes for better products. Josh Rathke, excellent work leading the team in this shift! • Onboarding Ian Schneider, Director of GTM at TryNow.com. Grit and Empathetic Leadership are two core value of ours at TryNow. Ian is as gritty as they come and leads with empathy. He's off to an electric start — 8 new deals created in week 1! ⚡ 🔥 • We're onboarding brands across a variety of new verticals (#TryNewThings): pajamas, fine jewelry, color cosmetics, streetwear, bedding, wedding dresses, and premium sandals. #TryBeforeYouBuy #TryNewThings #Shopify

31

2 Comments

There is a new #searchengine in town going after #Google and #Perplexity out of Canada A new AI-first search engine, Upend, has emerged from stealth, aiming to democratize access to advanced generative AI models. Upend is a Canadian startup that has just launched its AI-powered search platform, offering users a unique and powerful search experience. Access to 100+ Large Language Models (LLMs): Upend provides users with a wide selection of LLMs, including models from #OpenAI , #Claude , #Mistral , and more, allowing for diverse and specialized search capabilities. Upend offers the option to base search results on Wikipedia, providing users with grounded and cited information. Affordable Pricing: Upend aims to democratize access to advanced AI search by offering team plans starting at $20/month, significantly lower than competitors. Upend's mission to make AI search more accessible and affordable is a significant step towards empowering students, professionals, and businesses to harness the power of generative AI. _______ 🤩 Enjoy being the first to get news like this? ♻️ Repost it to your network and follow Ben Silverman for more! Want to stay up to date on all of the latest news and tools and get them directly to your inbox every week? ⛏️ Join our community of 2,000 + subscribers today https://lnkd.in/gMDQ6CDd https://lnkd.in/dneXZDbK https://lnkd.in/gfMvRbyq

1

I’ve noticed many founders don’t realize how much non-dilutive capital is flying under the radar in Canada. So I’ve shared a list of alternatives you should know about before choosing to raise capital: 1. Business Scale-up and Productivity (BSP): BSP provides funding to help you scale your operations and increase your market reach. It’s a great way to drive growth without sacrificing equity. 2. Industrial Research Assistance Program (IRAP): IRAP offers financial support for research and development. It’s perfect for taking your technical ideas to the next level. 3. Scientific Research and Experimental Development (SR&ED): SR&ED provides tax incentives for technical development. It’s one of the most generous programs in Canada, helping you reclaim a portion of your R&D costs. 4. FedDev Ontario: If you’re based in Southern Ontario, FedDev offers funding to help businesses scale, innovate, and compete globally. It’s a solid option for companies looking to make a big impact. 5. Canada Small Business Financing Program (CSBFP): This program helps small businesses access loans for purchasing or improving assets like equipment or property. It’s designed to help companies grow and invest in their infrastructure. 6. CanExport SMEs Program: If you’re looking to expand into international markets, CanExport provides funding to cover costs related to market research, business travel, and participation in international trade events. 7. Canada Media Fund (CMF): For those in the digital media industry, CMF offers funding for the development, production, and marketing of innovative media projects. 8. Ontario Centres of Excellence (OCE) Programs: OCE offers various funding programs aimed at driving innovation and commercialization in Ontario. Their programs often focus on collaboration between businesses and academic institutions. 9. Mitacs Accelerate: This program provides funding to support research internships, allowing companies to collaborate with universities to solve specific challenges while benefiting from academic expertise. 10. Strategic Innovation Fund (SIF): Targeted at larger-scale projects, SIF offers funding to support R&D, commercialization, and the growth of innovative businesses in Canada. 11. Women Entrepreneurship Strategy (WES) Ecosystem Fund: This fund supports organizations that help women entrepreneurs grow their businesses. It’s part of a broader effort to increase women’s participation in the economy. 12. Innovative Solutions Canada (ISC): This program helps Canadian startups by funding R&D projects that address government challenges. It’s a great way to secure funding while working on real-world problems. 13. Alberta Innovates: For businesses in Alberta, Alberta Innovates offers several funding programs to support R&D, commercialization, and business growth. They focus on fostering innovation in sectors like energy, health, and technology. Have I missed any?

16

3 Comments

Weekly VC deals done: Sixfold / USA (NYC) / $15M / Generative AI for risk analysis / Salesforce Ventures, Scale Venture Partners, Bessemer Venture Partners, Crystal Venture Partners Hoop / USA (NYC) / $5M / AI-powered task management / Index Ventures, Origin Ventures, Divergent Capital, Chingona Ventures Liminal / USA (Denver, CO) / $5M / Generative AI for data security / Fin Capital, High Alpha, Matchstick Ventures, Craft Ventures Scout Fund Signapse / UK (Guildford) / £2M / AI Sign Language translation software / Soulmates Ventures, Deeptech Seed Fund, Royal Association for Deaf people, Empirical Ventures, CEAS Investments, The FSE Group Tobiko Data / USA (San Mateo, CA) / $21.8M / Data infrastructure / Unusual Ventures, Theory Ventures, 20Sales, Fivetran CEO, Census CEO, MotherDuck CEO Taloflow / USA (San Francisco) / $1.3M / AI technology / Wonder Ventures, First Check Ventures SiTration / USA (Boston, MA) / $11.8M / Materials recovery / 2150, BHP Ventures, Extantia, Orion Industrial Ventures, Azolla Ventures, MIT E14 Fund Kanastra / Brazil (São Paulo) / $21M / Tech-driven back office for finance / Kaszek, Atlantico, Valor Capital, Quona Capital, QED Investors, Actyus HyperSpectral / USA (Alexandria, VA) / $8.5M / AI-powered spectral intelligence / RRE Ventures, Kibo Ventures, Correlation Ventures, GC&H Investments Advanced Medicine Partners / USA (Durham, NC) / $32M / Advanced medicines / Deerfield Management, ARCH Venture Partners Understory / USA (Madison, WI) / $15M / Insurance for climate-vulnerable industries / True Ventures, Prelude Ventures M0 / Switzerland (Zug) / $35M / Decentralized financial infrastructure / Bain Capital Crypto, Galaxy Ventures, Wintermute Ventures, GSR, Caladan, SCB 10X Wevr / USA (Los Angeles, CA) / $3.5M / Creative development and production / HTC Corp., Epic Games Dive Engineering Software / Germany (Berlin) / $10M / Simulation software / D. E. Shaw group, First Momentum Ventures, Segenia Capital, Senovo Capital Envisagenics / USA (NYC) / Undisclosed / RNA splicing therapeutics / Third Kind Venture Capital, Empire State Development, Red Cell Partners, Bristol Myers Squibb Testsigma / USA (San Francisco, CA) / $8.2M / AI-powered test automation platform / MassMutual Ventures, Accel, STRIVE, BoldCap Inventive / USA (San Francisco, CA) / $6.5M / Embedded AI for SaaS / Wing VC, Tokyo Black, angel investors group XREX / Taiwan (Taipei) / $18.75M / Blockchain-enabled financial service / Tether iCOMAT / UK (Bristol) / $22.5M / Advanced composite manufacturing / 8VC, NATO Innovation Fund, Solvay Ventures, Velocity Partners VC restor3d / USA (Durham, NC) / $70M / 3D printed orthopedic implants / Summers Value Partners, Trinity Capital Eko Health / USA (San Francisco, CA) / $41M / AI for early detection of diseases / ARTIS Ventures, Highland Capital Partners, NTTVC, Questa Capital, Double Point Ventures, EDBI, LG Technology Ventures Alinia / Spain (Barcelona) / $2.4M / Corporate alignment platform / Speedinvest, Precurs

4

Canada's VCs should also be on the map! @Phil Chen (Brex) for the inspiration (PamPam City is awesome!) Toronto: 36 VCs and counting. Who are we missing? BetaKit Siri Agrell Satish Kanwar BDC Georgian Inovia Capital Golden Ventures Panache Ventures Radical Ventures Brightspark Ventures OMERS Ventures MaRS Investment Accelerator Fund Relay Ventures Real Ventures Ripple Ventures Staircase Ventures Two Small Fish Ventures N49P Ventures Information Venture Partners Highline Beta Portage Round13 Capital Graphite Ventures GreenSky Ventures Whitecap Venture Partners Amplify Capital Good News Ventures StandUp Ventures GroundBreak Ventures Genesys Capital Plaza Ventures Sixty Degree Capital Impression Ventures Kensington Capital Partners Limited Lumira Ventures Alate Partners Storytime Capital First Ascent Ventures

279

79 Comments

Klaviyo is just an email tool for the Shopify ecosystem, right? Q2 earnings definitely tell a different story, and the company is quickly solidifying itself as a true fast-scaling public SaaS Co with a ton of room to run. The TLDR - stock was up 31% yesterday. Nuff said Quick Metrics Summary 🔢 Revenue of $222M ( ⬆ 35% YoY) Gross Margin: 79% (⬆ 4% since S1) - I'm shocked to see this holding so strong with increasing prevalence of SMS amongst the customer base which is an inherently lower margin product. Management sited less reliance on 3rd party technology as they roll their own solutions in email servers etc. Net Dollar Retention: 112% - continues to trend down, but they explain it away because of a big pricing shift they did in 2022 that has rolled out of the data Customers over $50k in ARR: 2,386 (⬆ 64% YoY) - hands down the most exciting thing about this business looking forward. They continue to crush their move up-market and are closing true enterprise customers like Samsonite. They also had a really interesting ancedote in the call about a grocery retailer they moved from a homegrown system onto Klaviyo. These are the type markets you don't think about but could be huge growth opps moving forward. Free Cash Flow Margin: 14% - generating significant cash and $700m+ on the balance sheet is very impressive Rule of 40: 50% ⭐ - continuing to trend down but still very healthy balance of both growth and profitability. Product and Strategy 🗺 It's always interesting to watch messaging oscilate quarter over quarter - a few quarters ago it was all Klaviyo One and CDP. Now, that emphasis is a bit more subdued and they're doing more story telling on what actual value/actions those products can drive. With a strong launch in France, they've proven internationalizing the product can lead to rapid growth in a market so I expect to see a lot more of that moving forward. Klaviyo is actually a really logical platform to leverage AI, and it seems like they're doing a lot of it across their product suite which is cool. Looking Forward 🔭 Klaviyo raised guidance on both revenue and profit and believe they'll continue to perform well in international markets as well as deepening penetration across CDP, SMS, and Reviews. It's been a wild rollercoaster for Klaviyo bouncing between $21-$39 over the last year since IPO, but I think as they continue to show the market they can reliably forecast and grow their business, that volatility should level out. #saas #earningsseason

62

2 Comments

Come join the Future of Frontier Technologies AI & blockchain event panel in Santa Monica! ✨ We are hosting an event on Friday, May 3rd during @Farcaster's FarCon weekend. Which will bring many blockchain experts & builders to LA! Emerging LA is collaborating with Lightspeed's Faction VC as a co-host, with Partner Timothy K. moderating a panel. The panel will discuss the questions that you are probably wondering about these technologies: • How can AI and blockchain work together? • How will these frontier technologies impact industries today? • What are the real-world use cases that these technologies solve? • What are the current dynamics between innovation and regulation? ​One on hand, AI offers the promise of abundance, democratizing content creation but will open the door to (deep)fakes and mistrust. On the other hand, ​Blockchain offers digital scarcity, authentication, and trust in a world inundated with digital replicas. --- The audience will be filled with founders, technical builders, investors, and industry executives in sectors that are most relevant to these trends: Gaming, media, content, finance, social. The early bird ticket sold out quickly last week and the general admission tickets expire on Wednesday, so don't miss out and grab yours soon! Partners for this event include Sunset Ventures and AI LA as well as others in town: • LA schools & clubs: Wharton Club of Southern California, Wharton Digital Assets Club, Harvard Kennedy School, Harvard Blockchain , and Blockchain at USC • CPG startups & Brands: Studio TBD, doorsa, Red Bull --- PROMO CODE DISCOUNTS! If you are a student or alum from Penn/Wharton, Harvard, USC, reach out and ask for code!! 😉 See you on Friday! (Link to event in comments) #ai #blockchain #farcon #web3 #LA

63

19 Comments

Consensus ideas held by most VCs on AI right now: 👉 Vertical AI agents are the solution to optimizing workflows in enterprises. 👉 LLMs are the fastest depreciating assets in history. 👉 Scaling laws will lead us to AGI (though they still invest in SaaS 🤔). What are non-consensus views, in your opinion? I’ll start: Service companies will primarily buy agents from hyperscalers and foundational LLM companies. This will especially enhance their capabilities in mission-critical services that require “humans in the loop.” Accounting firms, law firms, consulting firms, compliance, and tax advisory—service businesses are about to become very profitable.

10

1 Comment