Table of Contents
Skydive Investment Report
Category: AI agents / workplace automation
Company Stage: Newly launched product within a Series A-stage company
Founder or Founders: Dhruv Amin and Marcus Lowe
Headquarters: San Francisco, California
Funding: Parent company Anything raised an $11 million Series A at a reported $100 million valuation; CNBC also reports approximately $3 million of prior funding
Business Model: Subscription plus usage-based AI and compute charges; custom enterprise contracts
Product Hunt Launch Date: August 27, 2026
Report Date: August 30, 2026
| Investment Metric | Assessment |
|---|---|
| Venture Potential | 69/100 |
| Unicorn Path | Conditional |
| Valuation Attractiveness | Not Assessable |
| Evidence Confidence | 57/100 |
| Final Decision | DD |
Executive Summary
Skydive enables companies to create autonomous, role-specific AI agents that operate across business tools. Each agent receives a persistent cloud computer, memory, credentials, files and communication channels. Users can delegate work through the web, Slack, email, iMessage or a command-line interface rather than manually constructing a deterministic workflow (Skydive; Product Hunt).
The product is aimed initially at founders, operators and technology teams seeking to automate support, research, marketing, engineering operations and administrative work. Its strongest product signal is breadth: persistent environments, more than 600 claimed integrations, recurring routines, multi-agent collaboration, browser operation and team governance are already represented in the product and documentation. The frequent changelog also indicates unusually rapid execution.
The strongest company signal is the founding team. Amin and Lowe previously built the business to a reported $2.2 million run rate, deliberately replaced its original service model with AI software, and subsequently grew Anything to a company-reported $2 million annualized run rate within two weeks of relaunch. They have relevant product, engineering and company-building experience, although these commercial figures relate to Anything—not Skydive (CNBC).
The central concern is that Skydive-specific commercial validation is almost entirely unavailable. Revenue, active workspaces, paying customers, retention, workload success rates and gross margin have not been publicly disclosed. Product Hunt attention and founder-reported examples of agents operating inside Anything demonstrate interest and dogfooding, not external product-market fit.
Decision: DD. Skydive merits a founder meeting because the team, product scope, parent-company distribution and market timing could support a venture outcome. Investment should depend on verifying external paid usage, retention, autonomous-task reliability, unit economics and financing terms.
Product Overview
The customer problem is fragmented knowledge work: employees move information among email, Slack, browsers, CRMs, project-management systems and internal documents, while conventional automation requires users to specify and maintain every step.
Skydive’s alternative is a named agent assigned a responsibility. The agent can browse websites, use connected applications, maintain files and memory, execute scheduled routines, and communicate with users across multiple channels. Templates cover functions such as support, marketing, sales, engineering and operations (official website).
Current pricing is:
- Free: Small trial credit, up to two seats.
- Starter: $20 per month, up to five users and $20 of included usage credit.
- Team: $200 per month, unlimited users and $200 of included usage credit.
- Enterprise: Custom pricing, including SSO/SAML, audit logs and dedicated support.
- Dedicated iMessage lines cost $100 monthly per agent (pricing documentation).
Model tokens are charged at cost without markup; sandbox compute, search, email and other services also consume the balance (usage documentation). The product therefore replaces combinations of manual labor, chatbot prompting and workflow tools such as Zapier, Make and Gumloop.
Founder and Team Assessment
Dhruv Amin is co-founder and CEO of Anything. His public profile reports prior product roles at Google and YouTube, including work on YouTube TV, and computer-science and business education at Stanford (LinkedIn).
Marcus Lowe is co-founder and CEO, with prior Google product experience, engineering and product leadership at Resource before its acquisition by Gem, and an MIT computer-science background (LinkedIn). The combination provides credible product, technical and commercial capability.
Anything says its 15-person human team operates with more than 150 internal agents. This is company-reported, not independently audited, but represents meaningful dogfooding (launch announcement). The team is also actively developing the product, with weekly releases documented from April through July 2026 (changelog).
No previous founder exit has been verified. Skydive appears dependent on the same small team operating Anything’s existing app-builder business, creating prioritization and key-person risk.
Founder Assessment: Strong technical and product execution with demonstrated adaptation, but Skydive-specific commercial execution remains unproven.
Market Opportunity
The initial segment should be defined as technology-forward SMB and mid-market teams with approximately 10–250 employees, multiple cloud applications, and recurring support, operations, research or go-to-market workloads.
A reasonable analyst scenario is 250,000 potentially suitable organizations globally paying the current $2,400 annual Team price. That produces a $600 million annual base-subscription opportunity, before incremental usage and enterprise contracts:
250,000 organizations × $2,400 = $600 million
This is an analytical scenario, not a verified market count. Actual willingness to pay will depend on whether agents reliably replace measurable work rather than create supervision overhead.
Expansion opportunities include larger enterprises, external customer-facing agents, developer APIs, industry templates and international markets. The enterprise opportunity could be substantially larger if Skydive develops compliance, observability, evaluation and permission controls appropriate for sensitive workflows.
The market is large enough for venture-scale revenue, but not automatically attractive: Microsoft, Google, Salesforce, ServiceNow, OpenAI and established automation vendors already control major distribution surfaces.
Traction and Growth Signals
Skydive launched publicly on August 20, 2026 and on Product Hunt on August 27. Product Hunt showed roughly 780 followers and four reviews around the report date, but its ranking display was dynamic and inconsistent; these figures should be treated only as launch attention (Product Hunt; August 27 leaderboard).
Company-reported operational examples include 150 internal agents, an agent that allegedly helped 90 users publish apps in its first week, and agents handling support, engineering operations and sales. These examples are encouraging but have not been validated through customer references or workload logs (launch announcement).
The product has a strong release cadence and a live self-service offering. However, no reliable Skydive-specific information was found for revenue, paying workspaces, active users, retention, usage growth, customer concentration or gross margin. Public testimonials are few and mostly displayed by the company itself.
Traction Assessment: Strong launch activity and internal usage, but external commercial traction remains unverified.
Competitive Position
Direct competitors include Lindy, Relevance AI, Gumloop and Zapier Agents. Indirect alternatives include Make, n8n, Microsoft Copilot Studio, custom scripts, outsourced operators and ordinary employees.
Skydive’s present differentiation is its agent-centric interface: persistent identity, memory and isolated computer rather than a workflow canvas. Its low base pricing and unlimited-agent model may encourage experimentation. By comparison, Lindy charges per user and usage tier, while Relevance AI emphasizes enterprise workforce orchestration (Lindy pricing; Relevance AI pricing).
Switching costs could emerge through accumulated agent memory, custom skills, connected credentials and operating history. Nevertheless, foundational models and browser-control capabilities are available to many competitors. Skydive has not demonstrated proprietary data, network effects or a uniquely protected technical advantage.
There is also an inconsistency in Skydive’s own materials: a comparison page references a $99 Team plan and “flat” pricing, while the current pricing documentation lists $200 with metered usage. The current pricing page is the more authoritative source, but stale collateral creates avoidable customer confusion (comparison page; current pricing).
If a major platform launched equivalent functionality within six months, customers would stay only if Skydive proved materially better cross-platform execution, reliability, memory and ease of deployment. That advantage has not yet been established.
Defensibility Assessment: Low to Medium
Business Model and Economics
The model combines recurring workspace subscriptions, usage top-ups, iMessage add-ons and custom enterprise contracts. The current Team-plan ACV is $2,400 before extra usage.
Passing model tokens through at cost is customer-friendly but limits gross profit from token consumption. Gross margin depends on unused credits, subscription value above direct costs, compute pricing, enterprise software fees and operational efficiency. None is publicly quantified.
Autonomous browser agents may incur meaningful sandbox, inference, search and support costs. More usage does not necessarily increase gross profit proportionally if most spend is passed through. Skydive must demonstrate that enterprise platform and governance revenue grows faster than infrastructure expense.
The product can plausibly use product-led acquisition through the Anything user base, templates and team sharing. Enterprise sales, security reviews and implementation support will likely increase acquisition costs.
Unicorn Path
A 10× ARR multiple is assumed for a high-growth AI software company with strong retention and high software gross margins. Because pass-through usage revenue deserves a lower multiple, Skydive would need approximately $100 million of high-quality recurring revenue to support a $1 billion valuation:
$1 billion ÷ 10 = $100 million ARR
At the current $2,400 Team ACV, that requires approximately 41,700 paying Team workspaces. Alternatively, at a hypothetical blended $10,000 ACV, it requires 10,000 customers; at a hypothetical $50,000 enterprise ACV, it requires 2,000 enterprise customers. The latter figures are scenarios, not known pricing.
Reaching that scale requires repeatable distribution beyond Product Hunt, strong six- and twelve-month retention, enterprise governance, demonstrable labor ROI, reliable execution and gross margins that exclude or appropriately discount pass-through costs.
Unicorn Path: Conditional
Valuation Assessment
Anything’s September 2025 Series A was reported as $11 million at a $100 million valuation, led by Footwork with participation from M13 and prior investors Bessemer and Uncork (Anything announcement; CNBC).
Skydive is not publicly shown as a separately financed entity. Its current revenue, ownership allocation, fundraising status and current parent-company valuation are unknown.
Valuation Attractiveness: Not Assessable
Assessment requires current ARR by product, growth, gross margin, retention, burn, runway, round size, SAFE or priced-round terms, post-money valuation, liquidation preferences and investor ownership.
Key Risks
- No verified Skydive revenue, retention or paying-customer data.
- Agent errors could create financial, customer-service or production incidents.
- Broad access to email, code, browsers and local computers creates material security risk.
- Major software platforms can bundle similar agent capabilities.
- Low switching costs before meaningful memory and workflow history accumulate.
- Pass-through model pricing may constrain gross margin.
- Product reliability may deteriorate across long, ambiguous workflows.
- Small team is split across Skydive and Anything.
- SOC 2 Type I is weaker evidence than sustained Type II operating effectiveness (security page).
- Pricing and positioning inconsistencies may complicate customer trust.
Final Assessment
Venture Potential: 69/100
| Category | Score |
|---|---|
| Market Size and Expansion Potential | 17/20 |
| Traction and Growth Evidence | 7/20 |
| Founder and Team | 14/15 |
| Product Strength | 9/10 |
| Distribution Potential | 11/15 |
| Business Model and Economics | 6/10 |
| Defensibility | 5/10 |
| Total | 69/100 |
The strongest elements are the founders, product breadth, development velocity and access to Anything’s existing audience. The weakest are commercial validation, defensibility and unclear gross-margin economics.
Evidence Confidence: 57/100
Founder backgrounds, pricing, product activity, security claims and parent-company financing are reasonably documented. Internal deployment and user figures are company-reported. Market sizing and customer-count requirements are analyst scenarios. Skydive revenue, retention, customer count, margins, burn, runway and current financing terms remain unavailable.
Final Decision: DD
The company is strong enough for formal diligence, but not for an investment recommendation. The credible team and differentiated product justify investigation; the absence of Skydive-specific commercial data, uncertain economics and intense platform competition prevent a stronger decision.
Upgrade Conditions
- Verified Skydive ARR above $1 million with sustained monthly growth.
- At least 100 paying organizations outside Anything’s existing customer base.
- Six-month logo retention above 80%.
- Software gross margin above 70%, excluding pass-through model costs.
- Documented task-success rates and customer ROI.
- Repeatable acquisition outside founder audiences and Product Hunt.
- Enterprise references using multiple agents in production.
Downgrade Conditions
- Weak paid conversion or rapid usage decline after launch.
- High churn caused by unreliable task completion.
- Compute and support costs preventing attractive gross margins.
- Security incidents involving connected credentials or autonomous actions.
- Major platform replication without durable differentiation.
- Reduced product cadence or declining founder commitment.
Questions for Further Diligence
- What are current Skydive MRR, paying workspaces and revenue growth?
- How many weekly active agents and externally owned workspaces are there?
- What are 30-, 90- and 180-day workspace retention rates?
- What percentage of free workspaces convert to Starter, Team or Enterprise?
- What is gross margin after model, sandbox, search and support costs?
- What percentage of agent tasks complete without correction or human takeover?
- Which acquisition channels generate retained paying customers?
- How much Skydive usage comes from existing Anything users?
- What are burn, runway and team allocation between the two products?
- How are prompt injection, credential misuse and destructive actions controlled?
- What are the current cap table, financing valuation and round terms?
- What proprietary data or execution advantage should compound over the next two years?