AI Capital Rewards Capacity Control, Continuation Funds Go Mainstream, and Agentic Ops Reshape Venture Work
The gist
VC work is shifting from picking winners to managing capital access, liquidity engineering, and AI-assisted operating leverage inside the firm.
This week’s developments
AI Capital Is Rewarding Capacity Control, Not Just Product Demand
This week’s AI financings made the split explicit: Sierra raised $950M, Blitzy $200M, Panthalassa $140M, and DeepInfra $107M, with capital concentrating in infrastructure-adjacent and regulated enterprise businesses rather than broad app-layer exposure. DeepInfra and Panthalassa matter because the investment case rests on inference and compute access, not just software adoption. Nvidia-linked backstop structures are also reshaping infrastructure underwriting by shifting unsold-capacity risk onto Nvidia through minimum-revenue and GPU rental arrangements, making large datacenter and GPU builds easier to finance. For VC teams, the work is moving toward judging who has real capacity access, financeable supply, and enterprise-grade monetization.
How should we adjust sourcing to prioritize capacity control?
If you're an individual contributor
- App-layer hype is fading; capacity access is now the real edge.
- Get sharper on infra, inference, and enterprise monetization so you stay relevant when product demand alone no longer wins.
Sources
- Deterministic Infra for Non-Deterministic AI Agents - Nishant Gupta, Meta Superintelligence Labs — AI Engineer, June 29, 2026
Practical patterns for GPU efficiency, workload placement, and resilient orchestration for agentic AI systems.
- The Agentic Future (06.23.26): AI Inference Wars — 0xSammy: Decentralized AI, Robotics & Autonomous Systems, June 23, 2026
Explains the shifting economics of AI inference, including routing layers, open models, and cost-driven provider competition.
- Why Everyone Is Suddenly Talking About AI Factories — The TWIML AI Podcast with Sam Charrington, July 2, 2026
How to optimize token-level inference with GPUs, memory, routing, and power-aware architecture choices.
If you manage a team
- Your team must judge supply, not just demand, to spot real winners.
- Coach analysts to underwrite capacity control, financing structure, and enterprise-grade revenue quality—not just usage growth.
If you lead the organization
- Capital is favoring financeable capacity; your sourcing lens must shift.
- Reweight diligence toward GPU access, backstop structures, and monetization durability, or you'll back the wrong AI layer.
Sources
- The Data Center Valuation Model Breaks on the Compute Factory — Global Data Center Hub, July 1, 2026
Framework for pricing AI data centers using offtake, power economics, GPU cycles, and financing structures.
- Anjney Midha's Plan to Radically Lower the Price of Compute — Odd Lots, June 13, 2026
A leader’s view on forecasting demand, locking in capacity, and avoiding the risks of financialized compute speculation.
- Inside the AI Sprint, Understanding Anthropic's Strategy | Tomasz Tunguz, Theory Ventures — The Peel with Turner Novak, May 15, 2026
Explains GPU scarcity, infrastructure constraints, and operating tactics that determine which AI businesses can scale profitably.
Continuation Funds Are Becoming a Core VC Liquidity Tool
Venture and growth continuation volume surged from about $2.5 billion in 2023 to more than $9 billion in 2024, with Houlihan Lokey estimating global continuation fund volume at roughly $63 billion. That jump matters because continuation execution is moving from an occasional liquidity fix to a repeatable VC capability, with recognizable deals now including Lightspeed’s multi-asset fund for 10 late-stage assets and Inovia’s reported process drawing about $900 million of interest for a $334 million transaction. For practitioners, the skill set is increasingly operational: choosing which assets roll, defending valuation, structuring LP options, and managing governance and communication under scrutiny.
How should your team build continuation-fund capability now?
If you're an individual contributor
- Continuation deals are now a real skill path, not a niche side task.
- Learn asset selection, valuation defense, and LP communication—those judgment calls will separate you from pure execution support.
Sources
- A GP and LP unpack best (and worst) practices in co-investment — Private Equity Spotlight, June 3, 2026
Practical guidance on valuation, risk alignment, and upfront rules for better GP-LP co-investment decisions.
- Episode 46 - from Tashkent to Termsheet with Victor Orlovsky — The GoingVC Podcast, May 20, 2026
Practical guidance on sharing deal rationale, rejections, and process details to strengthen LP relationships.
If you manage a team
- Your team needs continuation-fund judgment, not just transaction support.
- Coach people on governance, LP options, and scrutiny handling; the team that can run these processes cleanly becomes indispensable.
Sources
- How to manage 3, 5, or 10 acquisitions at once without losing your mind — Midaxo Blog, June 26, 2026
Frameworks for standardizing workflows, ownership, reporting, and knowledge management across several simultaneous transactions.
- Founder Fridays No. 195 — Founder Fridays, May 22, 2026
Leadership, board consent, and diligence practices that help teams operate independently and avoid funding-round surprises.
If you lead the organization
- Continuation funds are becoming a core liquidity engine, not an exception.
- Build repeatable capability around asset selection, pricing, and LP management now, or you'll outsource a strategic lever to others.
Sources
- What’s next for value creation? — Private Equity Spotlight, May 26, 2026
How leaders align operating teams, incentives, and playbooks across partial exits, continuation funds, and growth initiatives.
- E398: Hamilton Lane ($1T AUM) on Venture Capital, AI, and Private Markets — How I Invest with David Weisburd, July 2, 2026
Explores LP management, DPI pressure, and why continuation vehicles are becoming a structural private-markets tool.
Agentic Infrastructure Replaces Manual Venture Operations
Meridian this week framed Scout AI as an end-to-end venture workflow layer spanning autonomous sector research, opportunity surfacing, LP-ready reporting, and a unified CRM, workflow, and portfolio monitoring stack across sourcing, diligence, IC management, and portfolio tracking over 26 million company records. Affinity kept pushing AI-native CRM features — AI Notetaker, Deal Assist, Automation Builder, and Affinity Sourcing — deeper into meeting capture, relationship management, sourcing, and workflow execution.
The operational evidence is already concrete. V7 reported a 35% productivity lift in the first month of automated diligence. Affinity and Bessemer said they reclaimed 234 analyst hours per year. BlackRock said AI-enabled research increased throughput to 10–15 companies per day from 2–3 manually, while data-room review is now being described as falling from days to hours or minutes. Early signs also suggest better screening and earlier portfolio alerts, not just faster admin.
For investors and platform teams, the job is shifting away from note-taking, spreadsheet upkeep, and memo assembly toward judgment: defining the right questions, checking AI outputs, and spending more time on exceptions, conviction, and founder relationships.
How should teams reallocate work as AI automates venture operations?
If you're an individual contributor
- Manual venture ops are fading; your edge is AI judgment, not admin.
- Learn to audit AI outputs, spot bad signals, and own founder-facing judgment—those skills will keep you indispensable.
Sources
- Inside Nathan's Second Brain: Daniel Miessler, Security Expert & Creator of PAI, Audits My AI Setup — Cognitive Revolution "How AI Changes Everything", May 30, 2026
A practical look at self-checking AI summaries, catching ambiguity, and quality-controlling generated work.
- Grading AI Fluency — The FishmanAF Newsletter, July 3, 2026
A practical reminder to challenge AI suggestions, catch errors, and iterate toward reliable analysis.
- Why Your AI Is Making You Busier: The 6-Part Framework for Real Delegation — Build to Thrive, July 1, 2026
Six-step checklist for building self-running AI workflows with triggers, execution, review gates, and learning loops.
If you manage a team
- Your team’s value is shifting from throughput to exception handling.
- Coach analysts to supervise AI, not just run process; reallocate time toward diligence judgment, sourcing quality, and coaching.
Sources
- DDQ: Celebrating 100 Episodes with Swimming with Allocators — Swimming with Allocators, June 3, 2026
Practical lessons on redesigning diligence, using AI cautiously, and moving referrals earlier to catch issues sooner.
- AI Bootcamp (Lesson 16): Schedule It, Refine It, and Retrospective — Latticework by MOI Global, May 29, 2026
Shows how to schedule, verify, and retrospect AI-driven investment research for durable team execution.
- Managing AI Agents at Scale Across BFSI Operations - with Yoav Naveh of Reindeer AI — The AI in Business Podcast, July 3, 2026
Governance model for supervising agentic AI, balancing compliance, accountability, and human review in operational teams.
If you lead the organization
- Your operating model still pays for work AI can now do faster.
- Redesign headcount and workflows around AI-first research, CRM, and portfolio monitoring; hire for judgment and systems thinking.
Sources
- The Claude Guide Every Founder Should Run Before Fundraising — The Founders Corner®, July 1, 2026
How top VCs use Claude to triage decks and what that means for diligence criteria and workflow design.
- Episode 834 | Eric Ries Revisits The Lean Startup and Discusses H... — Startups for the Rest of Us, May 26, 2026
Eric Ries explains how AI shifts tactical execution while durable startup principles still guide competitive strategy.
- Why $1B Exits are Dead — The a16z Show, May 29, 2026
How AI startups change venture risk, firm design, and portfolio strategy for finding rare winners.