Hi, it’s the HEC Paris Alternative Investments Society VC & Growth team.
Under the Hood is a bi-weekly newsletter about recent developments in the Venture Capital industry, our perspective on key trends, and deep dives into how VC works.
Today, we’re covering the Automation for Citizen Developers space with three sections:
Deal in the Spotlight - Prevision.io’s Series A and recent developments in the auto-AI space
Fantasy Portfolio - We believe in learning-by-doing. Our Fantasy Portfolio reflects our thinking on early-stage companies and includes investments that we would make with sufficient capital on hand. We start off today with a deep-dive on our first fantasy portfolio company in the AI-as-a-Service space, Abacus.ai
What intrigued us this week - Tips from industry insiders on breaking into VC
As valued members of our association, you receive this edition ahead of time and have the chance to influence future editions by providing your feedback and ideas!
🔦 Deal in the spotlight
(by Aarzoo Sharma)
Prevision.io raised a €6mn Series A round led by Upfront Ventures. Key facts about the deal:
Value Proposition: SaaS platform for any business user to create and use actionable AI
Founded: September 2016 / HQ: Paris
Stage: Series A / Funding: €6mn (€7.5mn total) / Estimated ARR: €2mn
Investors: Upfront, Net Ventures, SNCF Digital Ventures Fund (managed by HI INOV), Bpifrance
Prevision is on a mission to help businesses unlock the capabilities of AI through their auto-ML platform meant for data scientists, IT teams, and citizen developers (business users who do not have data/ coding expertise). Their platform uses client data and ML models to help users with task automation and forecasting across use cases. They do this via a hybrid business model that incorporates elements of SaaS and Marketplace businesses. The supply-side of the marketplace consists of a community of ‘makers’ (developers) and partners (cloud providers, consultants, and other tech partners).
🚀 Why you should really care about this deal
Prevision’s horizontal approach (the platform covers the entire app-development lifecycle and can cater to use cases across functions) is interesting in light of the rise of Vertical SaaS (Salesforce Ventures).
Their modular/ lego-block approach (they have an ‘App Store’ for models backed by a community of makers) likely lends a great deal of flexibility to how users can build and scale projects while cutting costs of development in the long run.
They’re offering ‘White Box AI’ which is key given the growing importance of explainability in AI models (here’s a16z's coverage on the topic).
Prevision encourages collaboration across teams while optimizing the experience for IT (with their functionality around migration, monitoring, documentation, and scalability) which probably help boost their acquisition and expansion metrics since IT teams often act as gatekeepers in enterprise sales.
🤔 What you need to believe in
Developing a fluid experience for business users. Prevision's long-term prospects hinge upon a wider-spread adoption in the customer organization. If the learning curve for citizen developers is too steep, the platform is at risk of being marginalized within the data science/ IT team. We'd like to see Prevision implement tools that aid communication between end-user and IT-department to make the set-up of models fluid rather than silo'ed, and provide options for citizen developers to set up models independently with crisp coaching and guidance from the platform.
Tackling data processing. Building and maintaining ML models require large amounts of data (check a16z’s take) and the knowledge to pre-process it - something that is comparatively easy to achieve in a developer-driven organization. The main target segment for Prevision’s solution, however, are those companies that either lack those capabilities or organizational set-up: traditional (read: non-tech) SMEs and large enterprises. Prevision will have a real shot at capturing significant market share if they provide value in the entire AI value chain and unlock data-related network effects through their platform.
💸 Fantasy Portfolio
(by Julius Müller)
Today, we are kick-starting our Fantasy Portfolio by adding auto-AI company Abacus.ai! Abacus.ai is democratizing access to sophisticated AI models so that every company can effortlessly embed cutting-edge AI into their applications.
Find our in-depth investment memo HERE.
🚀 Why we would invest:
Abacus.ai strikes a delicate balance between keeping the implementation of advanced AI-models easy enough for business users to champion them in the company while appealing to engineering experts with innovative approaches such as finding techniques of matching algorithms to customer data, automatically generating training sets with synthetic data, and researching ad-hoc debiasing techniques.
Their early traction with SMEs has shown that their business model works and they can successfully put sophisticated models into production - a task that a multitude of other AI-as-a-Service companies have failed on.
The founding team (ex-Google, Amazon, and Uber) and existing supporters (e.g. Index Ventures, & Eric Schmidt) are stellar and we believe they can scale this business to more businesses while keeping their innovative edge.
ℹ️ About our Fantasy Portfolio:
We believe that in order to become great VCs in the future, we should act as a VC even before it becomes a full-time job for any of us. Every addition to our fantasy portfolio will include an investment memo that lays out our reasoning. We will track our portfolio’s performance over time in this newsletter to see if we have what it takes to spot a good deal.
If you feel like we should take a deeper look at any start-up, feel free to share it with us!
📚 What intrigued us this week
Tips on breaking into VC by Pietro Invernizzi and Alexandre Dewez
Newsletters to keep up with recent deals Techcrunch, Axios, StrictlyVC
Newly rebranded HV Capital (formerly HV Holtzbrinck Ventures; Delivery Hero, SumUp, Sennder) is recruiting for internships in Berlin coming fresh off their €535M fundraising for fund VIII.
Thank you for reading! This version was written by Aarzoo Sharma and Julius Müller. Don’t hesitate to write to us at aarzoo.sharma@hec.edu or julius.muller@hec.edu to share feedback, resources, or to submit requests for future deep-dives!