Venturi is building the system-of-record for enterprise AI spend — an attribution layer that resolves every inference call to the team, service, cost center, and budget responsible. The wedge is that no SAP/Workday/Okta-equivalent exists yet for AI cost: data is fragmented across model billing, observability, finance, source control, and identity, and ownership is unclear. Richard McKinney surfaced the problem firsthand running cost-to-serve on public-cloud products at Salesforce.
As of the July 1 note, Venturi is raising $1M on a $12M post-money SAFE cap (Richard initially described it as uncapped, then corrected it to a $12M cap). Early commercial signals: a $749K two-year LOI with a $500M+ ARR enterprise (pending a security review that would convert it to a paid contract in August) and a $17.5K two-month paid pilot with a Series-B AI-native company beginning late July. Founder ran 144 customer-discovery interviews (Jan–mid-June) and reached YC's top 10% twice. At the July 1 note ~$80K was committed and Richard wanted to close in ~2 weeks — that timeline is now likely past; confirm current close status. This is a Miles-sourced pre-seed via a Solamon intro.
Cloud FinOps is a $15.8B market in 2026 heading to ~$27B by 2030 (Fortune Business Insights; MarketsandMarkets). The timing signal is sharp: AI cost management jumped from a concern of 31% of FinOps practitioners in 2024 to 98% in 2026, now the discipline's #1 priority, with ~29% of infrastructure budget estimated wasted (FinOps Foundation data, via axis-intelligence). Venturi's bet is that AI-specific, multi-system attribution is a new system-of-record category rather than a feature of existing cloud-cost tools.
| Player | Positioning | Stage | Edge vs. them |
|---|---|---|---|
| Venturi | Multi-system AI attribution: billing + finance + observability + repo + identity, in-VPC | Pre-Seed | — |
| CloudZero | Unit-economics + native AI cost tracking on OpenAI/Anthropic/Bedrock Web | Well-funded, Series C-stage | Venturi's differentiation is deeper cross-system attribution, not just model-cost tracking |
| Vantage | Self-serve FinOps, 25+ integrations, free tier Web | Series-funded | Vantage is mid-market/broad; Venturi targets deep enterprise attribution in-VPC |
| Finout / Cast AI / Kubecost | FinOps visibility & K8s cost Web | Series A/B + OSS | Cover cloud/K8s cost; none own end-to-end AI-inference attribution to budget owner |
Moat (to be proven): read-only in-VPC deployment plus deterministic + ML attribution across fragmented systems. The open question is whether that's durable versus a feature the hyperscalers or observability vendors ship natively.
Natural acquirers are observability and enterprise-software platforms adding first-party AI cost attribution — Datadog, ServiceNow, Snowflake, Salesforce — plus the hyperscalers (AWS, GCP, Azure). The FinOps-consolidation comp is IBM's $4.6B acquisition of Apptio (2023). At a $12M entry cap, a $200M+ outcome returns the round many times over, which is a plausible mid-case if Venturi becomes the attribution control plane.
A full founder-and-company assessment following the meeting: strong operator, real and timely problem, early commercial signals, but very early and technically ambitious. Recommendation in the note: continue diligence.
Richard is a Pitzer alum who grew up in foster care, became legal guardian to his younger brother at 18, and funded university by flipping cars bought in Tijuana and resold in California. He worked across Deloitte, PwC, Amazon, Cruise, and Salesforce — spanning consulting, technical product, security, privacy, and cost/accounting infrastructure. Most relevant prior experience: at Amazon he built the company's first global utility/carbon accounting system (processed $2.5B in utility transactions, enabled $257M in procurement savings in year one); at Cruise he was Chief of Staff to the VP of Security & Privacy; at Salesforce he was a Director of Product Management in security, later focused on cost-to-serve across public-cloud products, where he surfaced the AI cost-attribution problem that became Venturi. Master's in security from the University of Washington (part-time 2020–2023) and a one-year Oxford master's while working at Salesforce. Co-founder Gabriel Lowinger is also a Pitzer alum and his freshman-year roommate, with a Harvard PhD in bioinformatics/biostatistics and prior ML-scientist experience at NIH.
Company: Venturi is an "attribution layer for enterprise AI" — it ingests data from model billing, finance, observability, source control, and identity systems, then resolves AI inference calls to the team, service, cost center, project, and budget responsible. Buyers appear to be CIOs, CTOs, and FinOps leaders. There is no system of record for enterprise AI analogous to SAP/Workday/Okta; data is fragmented and ownership is unclear.
Stage/traction: pre-seed/early seed; 2 founders, hiring the first founding engineer; working product with a customer-deployment process and design-partner use, still early. Fundraising: raising $1M on a $12M SAFE; received $80K since the prior Wednesday and wanted to close in about two weeks. No closed recurring revenue. Signals: a nonbinding LOI of $749K over two years (contingent on a security review) with a $500M+ ARR company, and a $17.5K two-month paid pilot with an unnamed Series-B AI-native company beginning late July. 144 customer-discovery interviews from January through mid-June; applied to YC twice, top 10% both times, feedback to show more PMF. Product is read-only, deploys in the customer-owned VPC (data stays in the trust boundary), ~30-minute Terraform/CloudFormation onboarding. Recommendation: continue diligence — founder quality high, problem timely, early signals strong for the stage; open tensions are durable moat and whether Venturi is a must-have system of record vs. a point solution.
Follow-up email from Richard correcting the SAFE structure and laying out use of proceeds.
Really great meeting earlier, and thank you for taking the time. One correction: when you asked whether I had capped the SAFE, I said no, and that was my error. The round is a post-money SAFE with a $12M post-money valuation cap and no discount — the structure already built into the financial model. A capped SAFE is the correct instrument at this stage; I'd rather give you a clean number than leave it open.
On use of proceeds: the $1M funds founder runway to the seed, the core product and deployment work, and the security review that converts our signed LOI into a paid two-year contract in August, which is the most important near-term milestone. The full allocation and milestone gates are in the attached model. Mobile: 206-486-6162; financial model attached; I'll turn around your diligence questionnaire when you share it. Look forward to continuing later this week.
Intro from Solamon connected; Richard responds that Venturi has just opened its pre-seed and proposes call times.
Intro from Solamon went through for Richard McKinney; scheduling a call. Richard: "Miles, great to connect, and thanks for the note and the times. The timing works well — we've just opened our pre-seed round at Venturi, so I'd welcome the chance to walk you through what we're building. Given StoryHouse's enterprise-tech focus, there should be real substance here." He proposed Friday 9:30am–12:30pm PT or 11am–1:30pm PT Mon–Wed the following week, with July 15 at 10am as a fallback. Email: richard@venturi.systems; happy to share the deck ahead of time.