Veredas Partners is a two-GP micro-fund concentrating on physical AI, run by an operator (Voxel Field CTO, Harvey Mudd, CMU robotics) who has already co-invested alongside StoryHouse in Ultra Robotics and Sancho. Fund I ($4M, ~20 investments, fully deployed) produced one unicorn and paired the GPs with SH on two of our highest-conviction deep-tech names; Fund II sharpens the thesis to first-check exposure in physical AI — specifically "neo-integrators" like Ultra — while both GPs remain part-time operators. The wedge is technical sourcing plus a first-check position on Harvey-Mudd–adjacent robotics founders.
The relationship is early: one brief but positive call (2026-06-10), no deck reviewed, no formal pitch to the full IC. Vai has co-led one third-party priced round we can confirm (Hellbender $12.5M seed, May 2026, alongside Magarac Venture Partners) (Web). SH context: this would be a Fund II LP commitment sized as an ecosystem/diligence-partnership check — Josh and Scott are already soft-committed LPs (Internal). Miles's inclination is a small check, framed similarly to Josh Jones's $50K/1% position in SH Fund I.
Veredas is fundraising into the deepest physical-AI capital cycle on record: 2026 has already eclipsed the full 2025 total in VC deployment, driven by humanoid/robotics/AI-chip demand (Web). For an emerging manager writing $50–250K first checks, thesis-tightening to physical AI increases signal-to-noise but also puts Veredas into the most crowded emerging-manager lane in venture.
| Player | Positioning | Funding / Stage | Edge vs. them |
|---|---|---|---|
| Veredas Partners | First-check physical AI, operator-GPs, HMC-adjacent sourcing | Fund II raising, $15M target | — |
| Hyperstition Research | Kardashev-acceleration deep tech (robotics/space/energy/compute) | Fund I ~$4.1M, closing (Web) | Overlapping thesis, direct HMC peer; Veredas has Voxel/enterprise operator edge |
| 1517 Fund | Deep-tech pre-seed / Thiel-fellow adjacent | Fund V, $80M+ (Web) | 1517 has scale and brand; Veredas competes on technical intimacy at smaller entry point |
| Compound VC | Frontier-tech seed, technical GPs | Institutional multi-fund (Web) | Different check size and access lane; Veredas wins on Claremont sourcing |
| HMCINQ | Harvey Mudd student/alum accelerator | Accelerator-scale checks (Internal) | Complementary, not competitive; HMCINQ is upstream of Veredas |
Moat: Vai's Voxel operator role plus HMC + CMU-robotics network creates a durable technical-diligence and first-check lane that peer solo-GP funds cannot easily replicate, though the fund itself has no institutional franchise yet.
Return economics are LP-driven, not acquirer-driven. Emerging-manager Fund I medians land 1.3–1.8x TVPI, with 25% of sub-$350M funds clearing 2.5x (Web). Veredas Fund II at $15M is small enough that a single physical-AI winner — especially in the current M&A cycle (Nvidia–Groq $20B, Mobileye–Mentee $900M, Skild AI–Zebra Robotics all 2025–2026, Web) — can carry the fund; the more relevant SH question is whether the LP position unlocks diligence access and shared deal flow beyond what a quarterly relationship call already provides.
Brief but positive first working call with Vai. Framed Fund II as a physical-AI, first-check vehicle with sharper thesis than Fund I. Miles indicated interest in a Fund II pitch and shared deal flow.