For a company that started by helping machines see, V7 has developed a sharp eye for the next logical step. Its first product, V7 Darwin, is a training data engine that helps machine learning teams label medical images, factory floor video, and satellite photos faster and more accurately. It is a solid, unglamorous business, serving over 300 clients like GE Healthcare and Bayer [V7, November 2022]. But the founders, Alberto Rizzoli and Simon Edwardsson, noticed something else. Their customers, after building a computer vision model, would often ask for help automating the next step: the document-heavy, human-in-the-loop workflows that followed. So they built V7 Go, a platform that uses large language models to turn unstructured documents into structured data, promising transparency and, ambitiously, zero hallucinations [Perplexity Sonar Pro Brief].
V7 is now a company with two distinct engines. One feeds the AI of others. The other aims to replace busywork.
From pixels to paragraphs
The initial wedge was pure computer vision. In industries like healthcare and autonomous driving, training a model requires thousands of meticulously annotated images. A human might spend minutes circling a tumor in a DICOM scan or a pedestrian in a video frame. V7 Darwin uses AI to suggest those annotations, turning the human into a reviewer rather than a manual labeler. It is a productivity tool for the data science team, and it found a home in regulated, high-stakes fields where accuracy is non-negotiable [AWS Marketplace].
The expansion into V7 Go reads as a natural, if ambitious, adjacency. If you can reliably identify objects in a complex image, the logic goes, you should be able to reliably extract key terms from a 100-page private equity fund document. The Go platform connects to models from OpenAI, Anthropic, and Google, and lets users build multi-step AI agents for tasks like contract review, due diligence questionnaire completion, or extracting KPIs from a virtual data room [V7]. The sales pitch hinges on auditability: every AI-generated output is tied to a source highlight in the original document, creating a verifiable chain of custody for compliance officers.
The traction behind a $33 million Series A
In November 2022, V7 raised a $33 million Series A co-led by Radical Ventures and Temasek, which it called the largest Series A in its category by more than double [V7, November 2022]. The round came on the back of explosive, if from-a-small-base, growth metrics throughout that year. The company reported a 3x increase in annual recurring revenue, a 100x surge in organic website traffic, and a 5x expansion in team headcount [V7, November 2022]. That capital has fueled the build-out of V7 Go and an expansion of its client roster to over 350 companies, including Siemens, Merck, and MIT [Databricks, 2026].
The team, now estimated at 80-99 people [LinkedIn], is led by a founder with a long-term view on entrepreneurship. Alberto Rizzoli, the CEO, is a fourth-generation entrepreneur who started his first company, a 3D printing venture for education, at age 19 [LinkedIn]. He is a frequent voice on the AI conference circuit, from the VIEW Conference to the RAISE Summit, and writes extensively on productivity and automation [VIEW Conference, 2026]. Simon Edwardsson, the co-founder and CTO, provides the technical backbone [LinkedIn].
Two products, one integrated bet
Structurally, V7 is betting that its two products will speak to two different buyers inside the same enterprise, and perhaps eventually to each other. The table below outlines the split focus.
| Product | Core Function | Primary Buyer | Key Verticals |
|---|---|---|---|
| V7 Darwin | AI-assisted training data labeling for images & video | ML/Computer Vision teams | Healthcare, Manufacturing, Autonomous Driving, Life Sciences [AWS Marketplace] |
| V7 Go | Multimodal AI agent platform for document workflows | Finance, Legal, Insurance operations | Private Equity, Insurance, Legal [V7] |
The synergy is aspirational but clear. A manufacturer using Darwin to train a quality-control AI for its production line could later use Go to automate the analysis of supplier contracts or warranty claim documents. The company's narrative is about owning the entire pipeline from raw, unstructured data (pixels or text) to actionable, structured insight.
Where the wheels could come off
This two-track strategy is not without its execution risks. The market for AI data labeling is competitive and can be price-sensitive, while the market for document automation agents is becoming crowded with both startups and features from cloud hyperscalers. V7 must prove that its focus on transparency and auditability is a defensible moat, not just a feature. Furthermore, selling two different platforms to two different internal stakeholders,data scientists and business operations,requires a dual-track sales motion that can strain resources.
- Market dilution. The company must avoid being seen as a specialist in neither field. A medical imaging team might question commitment if resources shift toward finance AI agents, and vice versa.
- The integration gap. The promised synergy between Darwin and Go is currently more of a strategic slide than a shipped product feature. Tangible connectors that allow, for example, a Darwin-created dataset to automatically train a custom model for use in a Go agent, would make the story concrete.
- The incumbents' response. Large players like Scale AI for data labeling and the myriad of RPA and workflow automation giants are adding AI capabilities rapidly. V7's answer is its deep vertical integration in regulated sectors and its claim of superior accuracy and traceability.
The next twelve months
The key milestone for V7 will be demonstrating that V7 Go can achieve enterprise traction at a scale that matches or exceeds its Darwin business. Watch for named customer case studies in finance and insurance that move beyond pilot projects to multi-year, six-figure contracts. The company is also likely to pursue a Series B round in the near future to scale its go-to-market efforts for Go and deepen its vertical solutions.
Financially, the path is about unit economics. If a private equity analyst spends 20 hours a week manually pulling data from confidential information memorandums, and V7 Go can reduce that to 2 hours with verification, the value is clear. At a conservative loaded labor cost of $100 per hour, that's $1,800 of saved time per week, or over $90,000 annually. A platform that automates this for a team of ten analysts quickly justifies a serious software budget. The incumbent V7 must beat isn't another AI startup; it's the spreadsheet, the PDF reader, and the exhausted junior associate working past midnight. That's a battle fought not with model benchmarks, but with trust and time sheets.
Sources
- [V7, November 2022] V7 raises a $33m Series A to help teams build robust AI, faster | https://www.v7labs.com/news/v7-raises-33m-series-a
- [Perplexity Sonar Pro Brief] V7 company overview and product descriptions
- [AWS Marketplace] V7 Darwin on AWS Marketplace | https://aws.amazon.com/marketplace
- [Databricks, 2026] V7 customer mention | https://www.databricks.com
- [LinkedIn] V7 company page and founder profiles | https://uk.linkedin.com/company/v7labs
- [VIEW Conference, 2026] Alberto Rizzoli speaking engagement | https://viewconference.it
- [Fortune, April 2024] V7 Labs expands from data labeling into workplace automation | https://fortune.com/2024/04/10/v7-labs-v7-go-workplace-automation-data-labelling-llms-ai-agents/