Startup Project: Build the future

Nataraj speaks with Mike Collins, founder of Alumni Ventures, about how his firm opened up venture capital to individual investors through pooled funds and co-investing. They also dig into the structural problems in private markets, why companies stay private longer, and how AI is reshaping venture without eliminating the need for patience, judgment, and diversification.Key topics

• Mike Collins’ background in venture
• What Alumni Ventures is built to do
• How the alumni fund model works
• Why diversification matters in venture
• The long-term nature of venture investing
• Problems in today’s venture and public markets
• Why companies stay private longer
• AI, hype, and real innovation
• Why adoption is slower than demos
• How new venture firms get started
• Learning resources for retail investors

Timestamps

00:00 - Introduction and why Alumni Ventures matters

00:24 - Mike Collins’ VC background and founding Alumni Ventures

01:24 - The mission: access and education for individual investors

03:59 - How an alumni fund is structured and invested

05:55 - Why pooled capital and network scale matter
06:24 - Venture capital as a core engine of the economy
07:28 - Why most individuals need private-company exposure
08:51 - Why venture portfolios need diversification
10:33 - Green D fund leadership and how team sourcing works
11:52 - How deals get allocated across multiple funds
13:11 - Venture as a long-term, slow-compounding asset class
14:07 - Building a venture allocation over several years
15:04 - Why checking investments too often looks like trading
16:44 - The biggest accessibility problem in venture today
17:42 - Why public-company incentives have weakened
18:43 - Why companies delay going public
20:16 - Late-stage private companies and who captures the upside
21:14 - Why SpaceX illustrates the privatization of gains
22:51 - Public versus private markets and the role of competition
24:19 - AI as a real platform shift, not just hype
26:15 - Big opportunities beyond AI: energy, defense tech, healthcare
27:41 - The cultural bias toward doom and negative headlines
29:37 - Why staying private can help companies like Stripe
32:40 - How AI capital concentration affects the broader venture market
34:53 - The Series A squeeze and how market corrections happen
36:49 - The main paths to starting a new venture fund
39:15 - Why great companies still take decades to build
40:29 - Why AGI timelines are often faster in theory than in reality
42:24 - Regulatory backlash and the slower pace of adoption
43:18 - Self-driving cars as a cautionary example for AI timelines
44:54 - Why the last 5 percent of product adoption is the hardest
46:58 - AI as a tutor for learning venture capital
47:53 - Books and frameworks for understanding startups and VC
49:18 - How to filter noise and think in decades, not days
50:44 - Why the best investors often do the least trading
51:50 - Mike’s core investing rules and closing thoughts

Postman is the world's leading API platform, trusted by more than 40 million developers and 98% of the Fortune 500. In this episode, we trace how Postman went from a developer's side project into a global, AI-native enterprise platform. Abhinav Asthana shares how Postman reached profitability before raising venture capital, why developer products are so hard to monetize, why he refused to sell and moved from India to the US, and how AI agents are reshaping the company's next decade.

Key topics:

• Building Postman as a side project to solve API testing
• Reaching Ramen profitability with in-product upgrades
• Why developer products are hard to monetize
• Turning down acquisition and moving from India to the US
• Single-player tool to enterprise API platformBuilding in India vs. the US
• AI agents as a new class of API consumers
• Astro, Passport, Fabric, and Fern — the agent-native stack
• Thoughts on "token maxing" and AI spend

Timestamps:

00:00 - Introduction: Postman and Abhinav's journey

02:15 - Before Postman: becoming a developer in India

05:06 - BITS 360: the virtual campus tour that started it all

07:09 - Monetizing mobile and the rise of Instagram

08:32 - Parting ways and hacking on Postman on the side

09:08 - The original problem: testing APIs

11:05 - From side project to the main thing

11:58 - The first signals of demand (Chrome Web Store)

13:18 - Reaching Ramen profitability with upgrade packs

15:52 - Realizing Postman was a venture-scale business

17:46 - The enterprise journey and product validation

20:22 - Why he refused to sell Postman

22:10 - Moving to the US as an Indian founder

24:38 - What made Postman compound

27:01 - Why developers are hard to serve

28:32 - How vibe coding and agents change API consumption

35:02 - Building in India vs. the US

38:59 - How AI changed how Postman operates

41:22 - Thoughts on "token maxing"

43:55 - Inside Astro: an agentic operating system

45:17 - The agent identity Passport

47:08 - Where Postman goes in three to five years

48:27 - IPO plans?

Resources & Links:

Postman → https://www.postman.com/

Abhinav Asthana → https://www.linkedin.com/in/abhinavasthana/

Nataraj Sindam → https://www.linkedin.com/in/natarajsindam/

Startup Project Episodes → https://thestartupproject.io/episodes

Startup Project sits down with Sviat, CEO of Bright Machines, to unpack how the company is using software-first robotics to manufacture complex electronics closer to where they’re deployed. The conversation focuses on why AI infrastructure is a strategic category, how Bright Machines differs from traditional contract manufacturing, and what onshoring really means for speed, quality, and security.

Key Topics:

• In this episode, Sviat explains that Bright Machines is focused on AI infrastructure, specifically the electronics that go inside modern data centers, including compute nodes, storage, and racks.
• He traces the company’s thesis back to a broader idea: use software and robotics to manufacture electronics anywhere, then narrow that focus to the data center market as demand became clearer.
• The discussion breaks down the market stack, from chip designers like NVIDIA and AMD, to ODMs, OEMs, hyperscalers, and contract manufacturers.
• Sviat shares why data center hardware became the right bet before ChatGPT accelerated the market: the products are expensive, strategically important, and driven by quality and throughput more than labor cost alone.
• The show compares traditional assembly lines with Bright Machines’ approach, which uses more robotics, sensors, cameras, traceability, and humans in the loop where automation does not make sense.
• Sviat explains how Bright Machines starts with design, using Bright Designer to simulate and improve manufacturability before lines are built, which helps reduce bottlenecks and improve automation over time.
• He says the company’s main differentiator is its software platform, which orchestrates the line, powers smart skills for navigation and inspection, collects data, and feeds insights back into design.
• The conversation covers line flexibility, including how much can be reused when switching between CPU, GPU, or different accelerator-based server designs, and when end-of-arm tooling must change.
• Sviat says Bright Machines is growing rapidly, expects more than 3x growth this year, and can produce high volumes from a small number of sites because of robotics efficiency.
• The episode closes on the broader case for onshoring AI infrastructure manufacturing in the US: security, time to market, quality, and a labor shortage that makes robotics necessary.

Timestamps:06:39 - The market stack: chip designers, ODMs, OEMs, hyperscalers, and CMs

09:07 - Why Foxconn, Jabil, and similar contract manufacturers matter

10:04 - Why large factories still rely on massive manual labor

12:20 - Why data centers are different from cheap consumer electronics

13:49 - Security, strategic sectors, and why AI infrastructure belongs onshore

16:26 - The first Bright Machines product: CPU compute servers for a hyperscaler

17:58 - How the line works: modular stations, yields, and automation levels

19:26 - Bright Designer and design-for-manufacturing feedback loops

21:20 - Robots, sensors, traceability, and humans in the loop

22:19 - Why time to market matters as much as cost

23:31 - Yield and throughput: 98% line-level yields and up to 2x throughput

25:25 - The Bright Robotic Cell and how the assembly line is structured

27:35 - Reusability across products and when tooling changes are needed

30:31 - Manufacturing as a service, not repair or field service

31:24 - Growth, gigawatt-scale capacity, and output from a single site

33:00 - Why current hyperscaler capex is not expected to slow near term

34:45 - The bottlenecks before deployment: chips, components, power, permits

36:54 - Bright Machines’ three pillars: platform, data layer, and Bright Designer

39:15 - Why humanoid robotics is exciting but not ready for industrial use

41:16 - Where LLMs and newer AI tools can help the robotics workflow

43:57 - The overlooked advantages of onshoring manufacturing in the US

45:59 - What Bright Machines could build next: more complex electronics and future AI devices

Dive into the evolving role of AI in design and collaboration as Danny Wu, Head of AI Products at Canva, shares insights on how the platform is transforming creative workflows, democratizing design, and leveraging large language models and diffusion techniques.

In this episode:

• How Canva redefined abstraction layers in design, moving from pixel edits to object-based workflows
• The evolution of AI at Canva: from traditional ML to transformers and large language models
• The impact of ChatGPT integration on Canva's user experience and business growth
• Agentic AI: Canva’s approach to AI that acts as a collaborative partner in design
• Challenges and misconceptions about AI generative models in creative industries
• Future plans: video content tools and more AI-powered features

Timestamps:

• 00:00 - Introduction to Canva’s innovation in abstraction layers
• 01:12 - Danny Wu’s background and journey at Canva
• 02:46 - Transition from software engineer to Head of AI Products
• 04:25 - How diffusion and large language models accelerate Canva’s AI capabilities
• 06:50 - The dominance of transformer models in Canva’s AI strategy
• 08:51 - Shift from pixel to object, then conceptual design with AI
• 09:19 - Limitations of chat-based creativity vs direct manipulation
• 11:01 - The future of design involving AI-generated, editable content
• 13:30 - Launch of Canva AI and the platform's new architecture
• 15:22 - Use cases and limitations of AI in visual and video content
• 17:05 - Canva’s diverse user base and how AI personalization fits different needs
• 18:48 - Challenges of AI aesthetics and user customizations
• 20:32 - Amazing AI features like Magic Layers for editable image designs
• 22:16 - Comparing models: diffusion, open source, and proprietary tech
• 27:36 - Exploring agentic AI: Canva's vision of AI as a collaborative partner
• 30:37 - How ChatGPT and similar tools boost Canva’s reach and usability
• 34:26 - Personalizing AI output for users and reducing generated “cookie-cutter” content
• 37:38 - Managing AI's creative style and avoiding homogenization
• 42:39 - Canva’s feature development process and testing workflows
• 46:36 - Surprising use cases, like self-grading quizzes in education
• 48:59 - Overlap and differentiation between Canva and other design tools like Figma
• 50:54 - Future video tools and content creation enhancements at Canva

Explore how the latest advancements in AI are shifting from traditional training to inference-focused efficiencies, and how companies like Adaptation Labs are pioneering adaptive, full-stack AI solutions that democratize control across industries.

Key topics:

• The evolution from compute-heavy training models to efficient inference layers
• How inference costs are changing despite increasing AI demand
• The role of adaptive, gradient-free learning in democratizing AI customization
• Challenges with the last 5% reliability gap and continuous learning
• The importance of full-stack optimization—from data to interfaces in AI systems
• Future trends: decentralized AI, edge computing, and ongoing innovation

Timestamps:

• 00:00 - Introduction to AI trends: scaling vs inference efficiencies
• 01:01 - Sudip’s background: Google Brain, DeepMind, and inference infrastructure
• 01:34 - The rapid growth of foundation and large language models
• 02:36 - Comparing traditional ML project timelines to large foundation models
• 04:20 - The transformative potential of foundation models in enterprise and underserved communities
• 05:33 - The shift from task-specific models to general-purpose foundation models
• 07:07 - How inference costs have evolved: the rising demand vs falling per-token costs
• 08:37 - The challenge of inference in trillion-parameter models and the move towards smaller, verticalized models
• 10:14 - Factors driving high inference costs: model size, reasoning, agentic workloads
• 12:13 - The probabilistic nature of inference and API pricing complexities
• 13:07 - Variability in inference costs and demand in real-world scenarios
• 14:14 - The autoregressive, sequential nature of LLM inference and system challenges
• 16:45 - Cost implications of autoregressive inference and the move to more efficient, localized models
• 18:18 - The motivation behind Adaptation Labs: democratizing AI control and customization
• 19:47 - Adaptive, gradient-free continual learning and environment interaction
• 21:26 - Co-optimizing full-stack AI: systems, interfaces, and models
• 22:34 - How interface design impacts AI adoption and continuous learning
• 23:55 - The evolution of techniques: from foundational training to open-source innovations
• 26:18 - Handling the ‘last 5%’ reliability challenge in enterprise AI deployments
• 28:02 - The importance of system feedback and adaptive learning in coding and decision-making
• 31:12 - Adaptive Data and AutoScientist: seamless data transformation and model co-optimization
• 32:55 - Use cases: finance, low-resource languages, long context data
• 34:13 - The role of inference techniques and creating high-quality data for customization
• 36:10 - Future of adaptive, task-specific interfaces and continuous, real-time learning
• 38:49 - Full-stack AI: data, models, interfaces, and their iterative feedback loops
• 41:18 - The competition between fine-tuning and adaptive inference techniques
• 43:29 - The origin of new inference techniques: industry labs, open source, and innovation hubs
• 45:27 - The “last 5%” reliability gap: why it’s critical and how dynamic learning can help
• 48:27 - Hardware vs software optimization in AI systems and the future of systemic efficiency
• 51:25 - Growing AI demand, hardware constraints, and the opportunity for systemic innovation
• 52:48 - The shift from training to inference and decentralized AI models at the edge
• 54:12 - Final thoughts: the evolving landscape and long-term AI innovation

Connect with Sudip:

• LinkedIn

Connect with Nataraj:

• ⁠LinkedIn⁠

In this episode, we explore how Vast Data is revolutionizing storage solutions to support the exponential growth in AI workloads. Jeff Denworth shares insights into their innovative architecture, market strategy, competitive differentiation, and how they’re shaping enterprise data management in the era of AI.

Main topics:

• The origin and evolution of Vast Data’s innovative storage architecture since 2016
• How Vast’s solutions support large-scale AI and deep learning workloads
• The strategic focus on enterprise features, multi-tenancy, and integration with hyperscalers
• The impact of data reduction and cost efficiency on global flash supply
• New opportunities unlocked by Vast’s platform for analytics, vector search, and long context inference
• Business model nuances for cloud and on-premise deployments
• Vast’s profitability, market traction, and future growth prospects

Timestamps:00:00 - The AI super cycle and storage bottlenecks creating new opportunities02:20 - Understanding Vast Data's origin story and core architecture04:15 - How Google’s distributed systems influenced new storage innovations06:10 - Addressing scalability limitations of traditional storage systems08:00 - The shift from hard drives to flash and its market implications10:05 - Supporting AI workloads through scalable, enterprise-grade storage solutions12:00 - Customer sectors: life sciences, finance, and AI cloud providers14:15 - On-premise focus versus cloud deployment and hyperscaler strategies16:05 - Vast’s competitive differentiation: features, performance, and new data modalities18:15 - Integration with vector databases, analytics, and real-time AI inference20:30 - Business models: capacity-based, subscription, and partner collaborations22:50 - Addressing flash supply chain constraints and global market impact26:10 - The role of data reduction, federated data management, and long context storage30:50 - Unlocking enterprise data monetization and AI agent scalability34:15 - Impact of advanced storage on inference, context windows, and model efficiency36:50 - The current hardware procurement landscape and Vast’s software-led approach40:05 - Profitability metrics, growth, and the valuation of Vast Data42:25 - Final thoughts: the evolving data infrastructure landscape driving AI innovationResources & Links:Connect with Jeff Denworth:

In this episode, Garima Kapoor, co-founder and co-CEO of Min.io, shares insights into how storage infrastructure is evolving in response to AI, cloud, and enterprise needs. She offers a clear view of the market dynamics, innovative trends, and the strategic role of open-source technology in shaping the future.

Key topics:

• The origins and motivation behind Min.io’s development
• How data growth influences storage strategies and the shift toward hybrid and private clouds
• The impact of AI on storage infrastructure and workloads
• Competitive landscape with giants like AWS, Azure, GCP, and the rise of Neo Clouds
• The importance of open standards for application portability and data gravity
• Evolving customer adoption: from open source developer community to enterprise sales
• The role of AI in accelerating product development, coding, and organizational decision-making
• How AI’s rapid evolution is shifting the fundamentals of skills and fundamentals for engineers
• Future market opportunities: exponential growth in storage needs driven by AI and IoT

Timestamps:

• 00:00 - Introduction to Garima Kapoor and Min.io
• 00:31 - Motivation behind starting Min.io & market needs for object storage
• 01:07 - The founding story and personal drivers for creating Min.io
• 02:13 - Data growth drivers and the importance of data proximity over cloud location
• 03:05 - Business landscape: cloud vs. on-premises and hybrid environments
• 04:01 - Data migration challenges and promoting application portability
• 06:10 - Early product-market fit through open source and developer community growth
• 07:19 - Enterprise adoption journey from open source to cloud-native architecture
• 08:17 - Customer acquisition strategies blending bottom-up developer growth and enterprise sales
• 09:27 - Competing with Amazon, Microsoft, Google in the cloud storage space
• 11:33 - Impact of AI on storage: demand, infrastructure evolution, and market timing
• 12:51 - Min.io’s advantage in AI workloads due to cloud-native architecture
• 13:21 - Penetration of AI in storage: training, inferencing, and data utilization
• 15:01 - AI for enterprise applications: storage, models, and data lakes
• 16:26 - Neo Clouds and their role in GPU-optimized storage architectures
• 18:58 - The increasing demand for object storage driven by AI and data creation
• 21:02 - The effect of AI coding tools on product development speed and engineering skills
• 23:36 - Internal AI-driven solutions for operational efficiency
• 24:44 - The role of AI in reducing reliance on SaaS tools and infrastructure security
• 27:22 - Managing costs and building for the future in AI investment and storage
• 29:01 - The opportunity cost of tokens and AI-driven productivity gains
• 31:00 - Skills for early engineers in an AI-enabled future
• 33:32 - Min.io’s next steps and market expansion plans
• 34:36 - The paradigm shift: every business becoming AI and data-driven by 2026

Resources & Links:

Connect with Garima Kapoor:

• ⁠Min.io Official Website⁠

• ⁠Garima Kapoor - LinkedIn⁠
• ⁠OpenAI⁠
• ⁠NVIDIA GDC Announcements on Object Storage⁠
• ⁠Nataraj's previous interview on startup infrastructure⁠

• ⁠LinkedIn⁠
• ⁠Twitter⁠

Discover how Yutori is revolutionizing web interactions through autonomous AI agents designed for digital and web-based tasks. In this episode, Abhishek shares insights into building agentic AI, the technical challenges, and the evolving landscape of AI-powered automation.

Main insights:

• Yutori's founders come from Meta’s AI division, bringing top-tier expertise in AI and ML.
• The motivation behind Yutori's product stems from a long-standing interest in productivity tools and autonomous agents.
• Scouts by Yutori are AI agents monitoring web for specific signals, reducing manual browsing and keeping users up-to-date.
• The architecture relies heavily on specialized subagents, optimizing costs and relevance in web navigation.
• Abhishek emphasizes the transition from reactive to proactive AI, enabling agents to oversee tasks without constant prompts.
• The importance of user-centric design is reflected in a simplified UI, API integrations, and customizable workflows.
• Cost-effective strategies, like subagent architecture, help balance performance with scalability.
• The web is shrinking in terms of contribution and content creation; autonomous agents could change the landscape by managing and synthesizing information.
• Future product directions include deeper integrations, multi-task workflows, and enhanced proactivity in AI agents.
• Abhishek predicts a shift towards outcome-based pricing for AI tools, aligning value with costs.
• The conversation also explores implications for robotics, data generation, and the potential disruption of traditional content ecosystems.

Timestamps:

• 00:00 - Introduction to Yutori and its core product Scout
• 02:01 - Motivation for building autonomous AI agents
• 03:25 - The technical evolution from simulation to physical robots
• 04:44 - Origins of the Scout idea and focus on productivity tools
• 07:11 - The vision for web automation and agent-driven interactions
• 09:38 - Push vs Pull content systems and control over web consumption
• 10:38 - Demo of Scout setup and operation
• 14:00 - Technology foundation: web crawling, in-house navigation, and orchestration
• 16:28 - Data indexing and real-time monitoring approaches
• 18:20 - Subagents' distinct roles: navigator, researcher, social media scout
• 20:37 - Reporting, alerting, and workflows with Scout outputs
• 22:07 - Practical examples: monitoring market trends, personal tasks, and competitive intelligence
• 23:42 - Extending Scout functionality to actions and integrations
• 24:24 - The future vision: integrating Scout results into broader workflows
• 25:53 - Developer flexibility with subagents and API controls
• 27:03 - Cost considerations and architecture efficiencies
• 28:50 - The move towards proactive, autonomous agent behaviors
• 30:33 - Challenges of consumer adoption and simplifying interfaces
• 32:23 - Incentives for content creation and web ecosystem evolution
• 37:33 - Building trust and reliability in agent systems
• 39:18 - The web’s evolution and the rise of self-hosted content
• 41:24 - Impact of agent-based systems on content quality and SEO
• 43:32 - Measuring product-market fit and collecting user feedback
• 44:51 - Strategies for user acquisition and word-of-mouth growth
• 45:35 - Meta’s AI investments and industry trends
• 47:04 - Business models: subscription vs usage-based pricing
• 49:55 - Robotics advancements and synthetic data generation
• 52:22 - Final thoughts and opportunities for developers

Resources & Links:

• Yutori API → https://yutori.com/api
• Abhishek Das → https://abhishekdas.com/
• Nataraj Sindam → https://www.linkedin.com/in/natarajsindam/
• Startup Project Episodes → https://thestartupproject.io/episodes

In this episode, Varun, co-founder of Yoodli, shares insights into how his startup leverages AI to enhance communication skills, from public speaking to enterprise sales training. Tune in to understand how AI can empower humans rather than replace them, and the strategic evolution from consumer to enterprise products.

Key Topics:

• The origin story of Yoodli and its focus on helping people find their voice

• Transition from B2C to B2B: What was learned along the way

• The role of storytelling as a meta-skill in a world dominated by AI

• Using AI to make communication more authentic and human

• How large organizations like Google and Snowflake are integrating Yoodli

• The evolution of AI capabilities, from role plays to experiential learning

• Building modular, customizable AI products that adapt to customer needs

• The importance of deep integrations and the challenge of SaaS vendor proliferation

• Real-world growth stats: 900% revenue increase and millions of users

• Insights into leadership, authenticity on social media, and the value of vulnerability

• Personal stories from Sergey Brin’s projects and leadership lessons learned

Timestamps:

•  00:00 – Introduction to Varun and Yoodli’s journey

•  02:01 – Early days of Yoodli: Founding thesis and initial challenges

•  04:19 – Key lessons about public speaking skills

•  05:45 – The importance of recording and reviewing oneself

•  06:25 – Describing Yoodli as “Duolingo for public speaking”

•  07:25 – The role of storytelling in high-performance communication

•  08:21 – Building AI to enhance, not replace, human authenticity

•  09:07 – Judgment as a differentiator in AI-enabled work

•  10:01 – How Yoodli expanded into enterprise with Google & others

•  11:24 – Social media as a branding tool for founders

•  12:38 – The impact of authenticity on LinkedIn and lead generation

•  14:09 – The Google GTM training case study: How it started

•  15:07 – Product features for enterprise sales training

•  16:05 – Impact on sales onboarding and role play automation

•  17:32 – The future of experiential learning and AI role plays

•  20:17 – The broader vision for AI in education and training

•  21:26 – Impressive growth stats and customer insights

•  22:01 – The technological foundation: Modular AI architectures

•  23:52 – The influence of LLM improvements on product features

•  24:46 – The commoditization of AI role plays and experiential learning

•  25:12 – Building deep, customizable, scalable AI solutions

•  26:36 – The importance of scale and deep integrations

•  30:03 – Product differentiation through vertical focus and deep specialization

•  33:07 – Market challenges: Demand, consolidation, and customer expectations

•  34:42 – How to find and connect with Varun

•  35:30 – Sergey Brin’s projects, leadership lessons, and human insights

•  37:36 – Overcoming imposter syndrome: Everyone’s learning curve

• 39:01 – Final reflections and looking ahead

Resources & Links:

• Varun on Linkedin
• Nataraj on Linkedin
• Try Yoodli

Rethinking AI Compute Infrastructure: The TensorWave ApproachIn this episode, Jeff Tatarchuk, co-founder of TensorWave, shares how his deep industry experience and innovative mindset are transforming AI compute infrastructure. We explore how building specialized data centers, focusing on AMD GPUs, and creating flexible ecosystems are shaping the future of scalable AI.

In this episode:

• The evolution of cloud companies and the rise of Neo clouds focused on AI compute
• TensorWave’s unique strategy of deploying AMD GPUs in custom data centers
• Lessons learned from FPGA cloud business and transitioning into GPU infrastructure
• The technical challenges and solutions in scaling data centers quickly amidst power and supply chain constraints
• The importance of software ecosystems, interoperability, and supporting AMD’s software stack
• How TensorWave differentiates itself from purely financial arbitrage models and pure Nvidia-centric clouds
• AMD’s advantages in memory capacity, chiplet architecture, and software support
• The technical intricacies of CUDA versus ROCm, and efforts to build an open ecosystem
• Future vision: democratized, reliable, and flexible AI compute options for enterprise and labs

Timestamps:00:00 – Introduction to TensorWave and the AI compute landscape
02:30 – The rise of Neo clouds and innovation waves in cloud infrastructure
06:00 – How TensorWave’s FPGA cloud background shaped its GPU strategy
10:00 – Challenges in deploying large data centers: power, supply chain, and permitting
14:00 – Building and scaling AMD GPU data centers quickly and efficiently
19:00 – Software ecosystems: the CUDA moat and TensorWave’s ‘Beyond CUDA’ summit
23:00 – Market differentiation: technical and operational challenges in the Neo cloud space
27:00 – Supporting enterprise fine tuning and large-scale training demands
32:00 – AMD’s technical advantages: VRAM, chiplet architecture, and software support
36:00 – Building an open, heterogeneous AI ecosystem beyond CUDA
40:00 – What success looks like: a resilient, accessible AI compute future

Resources & Links:

• ⁠TensorWave⁠

• ⁠Beyond CUDA Summit⁠
• ⁠Scalar LM by Greg De Almos⁠
• ⁠AMD MI300X Data Center Chip⁠
• ⁠Nvidia H100⁠
• ⁠RoCM Software Stack⁠

• ⁠LinkedIn⁠
• ⁠Twitter⁠

This conversation offers a strategic look at how focused infrastructure development, software ecosystem support, and hardware differentiation are critical in shaping the future of accessible, scalable AI compute. Whether you're building data centers, developing AI hardware, or just interested in industry shifts, this episode provides valuable insights into how companies like TensorWave are reshaping the landscape.