Machine Minds

Venture capital looks glamorous from the outside, but the reality is far more nuanced. From surviving market cycles to backing founders through years of uncertainty, long-term success in venture comes down to judgment, grit, and pattern recognition earned the hard way. Peter Harris, Partner at University Growth Fund, brings a rare perspective shaped by nearly two decades in venture investing, student-led fund models, and firsthand experience navigating both booms and downturns in technology markets  .

Peter’s path into venture capital started early, influenced by entrepreneurship, real estate investing, and a shift from wanting to be an engineer to seeing business itself as a tool for solving problems at scale. After helping rebuild and operate one of the largest student-run venture funds in the country, he went on to co-found University Growth Fund, a diversified Series A and beyond firm with a mission that blends strong returns, student development, and economic impact.

The conversation spans what makes founders investable beyond pitch decks, why fundraising ability is often underestimated as a CEO skill, and how venture dynamics change when markets tighten. Peter also shares how AI is rapidly reshaping creation, distribution, and labor, and why ownership of assets may matter more than ever in the decade ahead.

Topics covered include:

• Peter’s journey from student venture investor to Partner at University Growth Fund and the lessons learned rebuilding a fund from the ground up
• How living through multiple market cycles changes how investors evaluate risk, founders, and timing
• Why a CEO’s ability to raise capital in both good and bad markets can determine a company’s survival
• The concept of grit in founders and why simply outlasting competitors can be a decisive advantage
• What Peter looks for beyond resumes including earned secrets, founder insight, and lived experience
• How University Growth Fund balances real venture execution with training the next generation of investors
• Why most businesses should not raise venture capital and the control trade-offs founders must accept if they do
• How AI is driving the cost of creation toward zero and shifting competitive advantage toward distribution, sales, and brand
• The implications of AI on labor markets and why asset ownership may become increasingly critical
• Common mistakes founders make when pitching VCs and how to think more clearly about what capital is actually needed for

For founders, operators, and anyone trying to understand how venture capital is evolving in an AI-driven world, this episode offers a grounded and experience-backed look at what really matters when building companies that last.

Connect with Peter Harris on LinkedIn: https://www.linkedin.com/in/vcpete/

Learn more about University Growth Fund: https://www.ugrowthfund.com/

Listen to Peter’s podcast, VC.fm: https://vc.fm/

Connect with Greg Toroosian on LinkedIn: https://www.linkedin.com/in/gregtoroosian/

Robotics does not stall because the ideas are bad. It stalls because the underlying infrastructure is missing. Adrian Macneil, co founder and CEO of Foxglove, has spent his career inside the systems that power some of the most ambitious autonomous technologies in the world, and he believes the next leap in robotics will not come from a single breakthrough robot, but from making robotics development radically easier for everyone.

Adrian’s path spans early work in payments and crypto, a formative chapter at Coinbase, and several pivotal years at Cruise during the early rise of self driving cars. At Cruise, he saw firsthand how much bespoke infrastructure was required to build, debug, and scale autonomy and how every leading AV company was quietly reinventing the same internal tooling. That realization became the foundation for Foxglove: a data and visualization platform designed to give robotics teams the same off the shelf leverage that software startups take for granted.

In this conversation, Greg and Adrian unpack:

• Adrian’s journey from early programming curiosity to building infrastructure at Coinbase and Cruise, and why autonomous vehicles made the value of robotics instantly tangible
• Why robotics development is dominated by custom tooling, siloed data formats, and painful debugging workflows, and how that slows the entire industry
• The origin of Foxglove and its mission to provide a shared data platform for robotics and physical AI, from logging and visualization to debugging and analysis
• What makes robotics data fundamentally different, including multimodal sensors, massive data volumes, limited bandwidth, and edge-first constraints
• The creation of MCAP as an open data format, and why interoperability is a prerequisite for robotics to scale beyond a handful of well funded teams
• How Foxglove acts as a single pane of glass for understanding robot behavior across simulations, incidents, and real world deployments
• Why robotics startups face “death by a thousand paper cuts,” from hardware and autonomy to go to market, pricing, and reliability expectations
• Lessons from fundraising in a non consensus market, and why finding investors who already believe your thesis matters more than convincing skeptics
• The parallels between today’s humanoid robotics hype and the early days of self driving cars, including the long tail of real world deployment
• What Foxglove looks for when hiring, and why proactive ownership is the mindset Adrian would clone across the entire company
• A ten year vision where starting a robotics company feels more like starting a SaaS company, with off the shelf infrastructure enabling founders to focus on real customer problems

If you care about the future of robotics, autonomy, and physical AI, and want to understand what actually needs to change for the industry to scale, this episode is a grounded and deeply informed look at the infrastructure beneath the hype.

Learn more about Foxglove: https://foxglove.dev

Connect with Adrian Macneil on LinkedIn: https://www.linkedin.com/in/adrianmacneil

Connect with Greg Toroosian on LinkedIn: https://www.linkedin.com/in/gregtoroosian

Robotics doesn’t fail in the field because of hardware alone—it fails when humans can’t understand, trust, or effectively work with the systems they’re given. Shakir Dzheyranov, founder and CEO of HelloRobo, has built his company around that reality.

With a background spanning visual arts, motion design, and product leadership at brands like Nike, Shakir brings a rare design-first lens to robotics and automation. After years in traditional design and marketing, he made a deliberate pivot into product design—drawn by the ability to tie design decisions directly to real user problems, business outcomes, and measurable impact. That shift ultimately led him to robotics, where he saw a massive gap between technical capability and human usability.

HelloRobo now operates as a specialized product design partner for robotics and automation companies, helping them build market-ready interfaces for robot operations, fleet management, and human–machine collaboration. Rather than chasing flashy MVPs or over-designed “vision concepts,” Shakir and his team focus on interfaces that can actually ship, scale, and be adopted by operators in the real world.

In this conversation, Greg and Shakir dive into:

• Shakir’s journey from art direction and brand design to building a robotics-focused product design firm
• Why robotics companies struggle with UX—and why established design patterns often don’t exist yet
• The decision to narrow HelloRobo’s focus exclusively to robotics and automation
• How contributing to Open Robotics helped establish credibility before commercial traction
• What makes UX in robotics fundamentally different from consumer software
• Designing operator interfaces and fleet management systems for complex, safety-critical environments
• Lessons from working with Bedrock Robotics, including designing new interaction patterns from scratch
• Why talking directly to operators beats copying existing UI patterns
• The three traits HelloRobo looks for when hiring designers: product thinking, visual clarity, and the ability to embrace chaos
• How design teams can stay grounded in business metrics instead of aesthetics alone
• The difference between MVPs, overbuilt “vision” products, and what Shakir calls market-ready software
• Why onboarding and education are still missing pieces in most robotics products
• Building culture in a creative, distributed team—and why HelloRobo is opening its first New York office
• Founder lessons on risk, playfulness, and learning to build without over-controlling outcomes
• What excites Shakir most about the future of robotics: mobility, prosthetics, and technologies that extend human capability

For founders building robotic systems, leaders scaling hardware companies, or anyone thinking about how humans actually interact with autonomous machines, this episode is a reminder that great robotics isn’t just engineered—it’s designed.

Learn more about HelloRobo: https://hellorobo.co

Connect with Shakir Dzheyranov on LinkedIn: https://www.linkedin.com/in/shakir-works

Connect with Greg Toroosian on LinkedIn: https://www.linkedin.com/in/gregtoroosian

In this episode of Machine Minds, we step beyond today’s transformer-dominated AI landscape and into a deeper conversation about what’s missing on the path to truly autonomous, long-horizon intelligence. Jacob Buckman, co-founder and CEO of Manifest AI, joins Greg to explore why current AI systems struggle with long-term reasoning, persistent memory, and extended task execution—and what it will take to unlock the next paradigm.

Jacob’s journey into AI began early, fueled by science fiction, programming, and a fascination with building systems that could do meaningful work autonomously. From studying and conducting research at Carnegie Mellon to working at Google Brain, he watched deep learning unify once-fragmented AI subfields—vision, language, speech—under a single scalable framework. That unification shaped his conviction that the next breakthrough wouldn’t come from incremental tuning, but from rethinking a fundamental architectural bottleneck.

At Manifest AI, Jacob and his team are tackling what they believe is the missing piece: scalable long-context intelligence. Their work centers on replacing transformer attention with a new family of architectures called retention models, designed to compress and retain relevant information over time—rather than repeatedly replaying massive histories. The goal: AI systems that can reason, learn, and work continuously over hours, days, or longer.

In this conversation, Greg and Jacob explore:

• Jacob’s path from aspiring scientist to AI researcher and founder—and why curiosity plus first principles thinking matter more than trends
• Why today’s large language models excel at short tasks but break down over long horizons
• The core limitation of transformer attention—and why “attention is all you need” may no longer hold
• How retention architectures unify the strengths of transformers and recurrent neural networks
• What it means for an AI system to compress knowledge instead of endlessly appending memory
• Why long-term reasoning, iterative problem solving, and true autonomy require architectural change—not orchestration hacks
• The misconception that agent orchestration can substitute for unified, persistent intelligence
• How long-context models could reshape agents from short-lived “consultants” into persistent, personalized collaborators
• The technical challenge of translating theoretical breakthroughs into high-performance GPU kernels
• Why Manifest AI is open source—and how their work aims to move the entire field forward
• Lessons from unifying AI subfields, the “bitter lesson” of scale, and avoiding ad-hoc solutions that won’t last
• Jacob’s view on cost, intelligence density, and why better architectures will increase—not reduce—investment in AI
• Advice for founders and researchers: focus relentlessly on the single bottleneck that matters most

If you’re building AI systems, researching foundations of intelligence, or trying to understand what comes after today’s models, this episode offers a rare, deeply reasoned look at where the field may be heading—and why architectural simplicity could unlock far more than brute force scale.

Learn more about Manifest AI: https://manifestai.com

Explore the open-source retention models: pip install retention

Connect with Jacob Buckman on LinkedIn: https://www.linkedin.com/in/jacobbuckman

Connect with Greg Toroosian on Linkedin: https://www.linkedin.com/in/gregtoroosian

In this episode of Machine Minds, we dive into one of the most overlooked choke points in logistics: the loading dock. Chris Smith, founder and CEO of Slip Robotics, joins Greg to unpack why loading and unloading trucks remains one of the most manual, time-consuming processes in modern supply chains—and how Slip is transforming it with autonomous, high-payload mobile robots.

Chris brings a rare blend of firsthand operator insight and deep robotics experience. From continuous improvement roles at Cummins, to factory-scale automation at Tesla, to building heavy-payload robots at a previous startup, his career repeatedly exposed the same inefficiency: trucks spending hours being loaded and unloaded for trips that last minutes. Slip Robotics was born from the realization that this problem wasn’t unsolved because it was hard—it was unsolved because no one had redesigned the workflow end-to-end.

Slip’s solution is deceptively simple: a large, flat autonomous robot that becomes the staging lane, drives directly into trailers, stays with the freight during transport, and unloads autonomously on the other side. The result is five-minute loads, five-minute unloads, reduced labor, faster throughput, and dramatically improved dock utilization—without modifying trailers or docks.

In this conversation, Greg and Chris explore:

• Chris’s path from risk-averse engineer to founder—and why betting on yourself can be the most rational risk
• The real-world pain that inspired Slip Robotics, observed firsthand across manufacturing, warehousing, and factory logistics
• Why docks function as the “physical API” of logistics—and how Slip leverages that instead of fighting it
• The technical challenges of moving 12,000-pound payloads up dock levelers and trailer slopes
• Why Slip’s robots generate massive ROI despite moving only ~5% of the time
• Early customer deployments, including automotive OEMs cutting lead times in half within weeks
• How close, on-site collaboration with early customers shaped a scalable, commercial product
• Why Slip operates on a Robotics-as-a-Service (RaaS) model—and how it aligns incentives around uptime and performance
• Hiring for early-stage robotics startups: why the first 20–30 people must be both specialists and adaptable
• Slip’s core cultural principles, including “rapidly deploying elegantly simple solutions” and delivering disproportionate value to stakeholders
• Where logistics automation is headed—and why dock efficiency may be the biggest unlock in transportation networks
• Chris’s take on general-purpose robots vs. highly specialized form factors, and where humanoids may (or may not) fit

If you’re building robots, deploying automation, or thinking about how physical systems actually scale in real-world environments, this episode offers a clear-eyed look at how simplicity, speed, and workflow-first thinking can unlock massive value.

Learn more about Slip Robotics: https://sliprobotics.com

Connect with Chris Smith on LinkedIn: https://www.linkedin.com/in/christopher-r-smith

Connect with Greg Toroosian on LinkedIn: https://www.linkedin.com/in/gregtoroosian

In this episode of Machine Minds, we look at how physical intelligence—the fusion of robotics, automation, and software—can reshape one of society’s most strained systems: healthcare. Director of Software Engineering Nicholas Kirsch joins Greg to break down why hospital pharmacies are essentially “mini warehouses,” how automation is already quietly at work behind the scenes, and what it will take to reach the vision of a fully autonomous pharmacy.

Nicholas brings a rare dual perspective: a mechanical engineer turned software leader who spent years in Pittsburgh’s startup ecosystem building mobile manipulation systems, AMRs, and government-funded robotics programs before shifting into healthcare tech. His experience—from garage-stage startups to acquisitions and rebrands—gives him a clear lens on what it takes to scale robots from impressive demos to mission-critical reliability. At Omnicell, he now helps drive software for medication-picking systems, IV-compounding robots, and the next wave of automation designed to return pharmacists and clinicians to the work they trained for: caring for patients.

In this conversation, Greg and Nicholas explore:

• Why hospital pharmacies operate like 24/7 logistics centers—and why automation is overdue
• The long, largely unseen history of medication-picking robots (30 years and counting)
• What the autonomous pharmacy roadmap looks like, and why most hospitals are still at level 1 or 2
• The hard truth about robotics in healthcare: reliability isn’t a target, it’s a requirement
• How systems like Omnicell’s IV compounding and XR2 picking platforms reduce waste, increase traceability, and free clinicians from manual labor
• Lessons from Nicholas’s journey through multiple robotics companies, acquisitions, and pivots—and how software talent evolves within physical systems
• What he looks for when hiring software engineers in mission-critical environments, including curiosity, culture fit, and growth mindset over rigid credentials
• The promise (and limits) of AI in physical automation, and why general physical intelligence will unlock far more than humanoids

For anyone building automation in regulated environments—or simply trying to understand how robotics can meaningfully improve patient care—this episode offers a grounded, insightful look at the future of healthcare efficiency.

Connect with Nicholas on Linkedin: https://www.linkedin.com/in/nicholaskirsch

Connect with Greg on Linkedin: https://www.linkedin.com/in/gregtoroosian

In this episode of Machine Minds, we explore what it really takes to build the human infrastructure behind the coming wave of robots. CEO and co-founder Richard Petrazzini joins Greg to unpack how “robotic workforce enablement” can make or break uptime, customer trust, and the long-term success of robotics deployments—especially as robots leave cages, connect to the cloud, and move into human environments.

Richard brings an unusually layered background to robotics: he’s a third-generation biochemist who grew up watching labs evolve from manual rabbit and frog testing to fully automated, 10,000-test-per-day facilities. That firsthand view of how healthcare automated over decades now informs how he thinks about robotics support, standards, and scale. After building one of Argentina’s largest labs and co-founding a successful software company, he spent two years doing nothing but researching robotics before launching Robotic Crew, a nearshore “human in the loop” partner for robot fleets.

In this conversation, Greg and Richard get into:

• Why robotics is entering its “lab automation moment” – and what other industries can learn from decades of healthcare and manufacturing automation
• How Robotic Crew supports pioneers deploying robots in uncharted territory, from first unit to scaled fleets
• The logic of nearshore support for robots: Latin American talent, daylight operations, and why robotics can’t be treated like asynchronous software outsourcing
• “Robotears,” tiered support (Tier 1–3 and beyond), and how to build career paths for people who keep robots productive in the field
• The trade-offs founders face around uptime, ownership of incident data, and when to bring in a partner for support
• What it feels like when a startup hits real momentum—not hype—and why Richard believes robotics companies will run out of engineers long before they run out of demand
• A real-world case study improving response and resolution times for autonomous robot kitchens, and what that reveals about process, telemetry, and customer success in robotics

If you’re deploying robots, building them, or thinking about how to support a growing fleet without burning out your core team, this episode is a deep dive into the future of robotic workforce enablement—and the humans who will keep that future running.

Connect with Richard on Linkedin: https://www.linkedin.com/in/rpetrazzini

Connect with Greg on Linkedin: https://www.linkedin.com/in/gregtoroosian/

In this episode of Machine Minds, we look at what really happens when hardware innovation meets business strategy in robotics. Director of Business Development Milt Walker joins Greg to unpack how manufacturers, startups, and ecosystem players can scale robotics responsibly—without getting stuck in proof-of-concept purgatory or pretending they’re “just” software companies. From functional safety to workforce gaps and reshoring, Milt explores the forces reshaping how robots get built, deployed, and trusted in the real world.

Milt brings a rare blend of hands-on engineering, ecosystem building, and strategic thinking. From early days fixing XT and 386 machines in a small-town computer shop, through field roles and leadership at Intel and Sick, to helping seed the Atlanta robotics community with RoboGeorgia, he’s consistently sat at the intersection of technology and partnerships. Now at Next Cobot, he focuses on giving robot builders the safety-critical and control platforms they need so they can innovate faster on what makes their systems unique.

In this conversation, Greg and Milt explore:

• Why “alignment” is Milt’s non-negotiable in partnerships—and how it changes the BD conversation
• The shift from customization to configuration as a path to real scalability
• How mid-market manufacturers and smaller warehouses can approach automation differently from the top 10% of global players
• The growing importance of safety platforms and standards as humanoids and advanced AI move closer to the factory floor
• What it actually looks like to build bridges between Taiwan and US robotics ecosystems
• Milt’s moonshot fascination with nano-robotics and “bottom-up” fabrication
• Practical career advice for engineers and technical leaders who want to move into business development and strategy roles in robotics

If you’re building robots, selling them, or betting your operations on them, this episode will sharpen how you think about partners, platforms, and the long game of automation.

Connect with Milt Walker on Linkedin: https://www.linkedin.com/in/mwalkeriii/
Connect with Greg Toroosian on Linkedin: https://www.linkedin.com/in/gregtoroosian/

Chapters:
2:34 Essential Technology
3:31 Early Career Insights
4:42 Robotics Passion
7:38 Business Development Path
9:58 Atlanta's Robotics Ecosystem
12:37 Joining NexCobot
15:19 Partnership Strategies
16:14 Market Opportunities
18:33 Barriers to Adoption
21:35 Finding the Right Partners
24:15 Successful Collaborations
27:07 Customization vs. Scalability
30:44 Technological Trends
37:07 Standards and Regulations
45:41 Moonshot Ideas
47:06 Advice for Aspiring Professionals
49:21 Reflections on Career Growth

What does it take to guide a robotics startup from a napkin sketch to a $775 million exit? Oliver Mitchell, a venture capital partner at FF Venture Capital and author of "A Startup Field Guide in the Age of Robotics and AI" returns to Machine Minds to share hard-won lessons from the trenches of deep tech investing. From the dramatic rise and fall of Webvan to the triumph of Kiva Systems (now Amazon Robotics), Oliver reveals why product-market fit isn't found in the lab - it's discovered in relentless customer conversations. We explore the pivotal mistakes founders make, why pivoting isn't failure but strategy, and how the right cap table can be the difference between scaling and stalling out.

Highlights:

• The Webvan lesson: How a $800M IPO collapsed because marketing never asked customers what they actually wanted... and how Kiva Systems learned from that failure to build the robots that became Amazon Robotics.
• The 500-customer rule: Why Oliver insists founders speak to hundreds of potential customers before building, and how CIV Robotics pivoted from drones to terrestrial robots by listening to construction giants like Bechtel.
• Five rules for automation startups: High utilization, 99.9% reliability, clear payback periods, defensible IP, and knowing your low-hanging fruit—Oliver breaks down what separates successful hardware companies from the graveyard.
• Culture as a competitive moat: Drawing from Netflix's "keeper test" and Pixar's collaborative creative process, Oliver explains why hiring one superstar beats five mediocre players - and why arrogance has no place in a startup.
• Cap table strategy: Why raising at an inflated valuation can doom your next round, how to price for competitive interest, and when to choose government grants, corporate VCs, or traditional funds.
• Defense tech is heating up: Oliver's new focus on Israeli technologies emerging from elite military units, the Golden Dome initiative, and why Apptronik's humanoid robots are built contract-first for shipbuilding - not household chores.

Oliver's book "A Startup Field Guide in the Age of Robotics and AI" is available on Amazon or at routledge.com (20% off with discount code 25ESA3 - valid 1 July 2025 - 31 December 2025)

Connect with Oliver Mitchell on Linkedin: https://www.linkedin.com/in/oliverbmitchell/
Connect with Greg on Linkedin: https://www.linkedin.com/in/gregtoroosian/

Construction is one of the world’s most complex, risk-filled, and operationally fragmented industries — and also one of the last to benefit from automation. At Raise Robotics, CTO Rishabh Aggarwal is helping change that.

From capturing diesel soot to turn into ink, to designing CubeSats at NASA, to building autonomous harvesting robots in ag-tech, Rishabh’s journey has always been about one thing: building real technology that solves real problems.  

In this episode, he breaks down how Raise Robotics is building a multi-use, autonomous construction robot capable of performing high-precision tasks on the edge of skyscrapers — layout marking, drilling, inspection, and more — all with 1/16-inch accuracy on chaotic, unpredictable job sites.  

We explore the technical challenges, the industry dynamics, the hiring realities of early-stage robotics, and the future of “physical AI” as robotics collides with deep learning at industrial scale.

⸻

Highlights
•From building off-road race cars to capturing diesel pollution:
Rishabh’s early startup experience involved inventing a device that trapped generator soot and converted it into ink — a hands-on lesson in iterative engineering and impact-driven design.  
•Learning hard lessons in hardware:
While building agricultural robots, a battery fire caused a full site evacuation — shaping Rishabh’s philosophy that startups should buy first, innovate later until product-market fit is clear.  
•Why construction needed robotics yesterday:
Multi-million-dollar rework, misaligned facades, and error-prone layout marking inspired Raise Robotics to automate tasks that require extreme accuracy — tasks humans aren’t well-suited for on 2,000-foot floors.  
•Inside Raise’s autonomous platform:
A robot on wheels equipped with long-reach arms, tool-changing capabilities, and precise localization — delivering layout marking, drilling, inspection, and more, while generating detailed QC reports for contractors.  
•The toughest technical challenge:
Designing a robotic arm that remains rock-solid on the literal edge of a building while working with sub-inch tolerance — in rain, dust, debris, and unpredictable environments.  
•Building a high-performing robotics team:
Rishabh emphasizes hiring for intent, attitude, and intrinsic motivation, not just technical skill — and why people from big tech increasingly seek startup impact.  
•Where robotics is headed next:
The rise of “physical AI,” the explosion of better sensors and compute, and the coming wave of human-robot collaboration that replaces cages with perception-driven safety.  
•Long-term vision:
A multi-industry platform that moves beyond construction into manufacturing, oil & gas, and high-risk industrial environments — wherever a precise, ruggedized robotic teammate is needed.  

⸻

Learn more about Raise Robotics:
https://raiserobotics.ai/

Connect with Rishabh Aggarwal:
https://www.linkedin.com/in/rish-creator/

Follow the Raise Robotics team on LinkedIn (they post regular updates):
https://www.linkedin.com/company/raise-robotics/

Connect with Greg:
https://www.linkedin.com/in/gregtoroosian/