The Venture Codex Logo

The Venture Codex

43 Ventures

1070 Maryland Street, San Francisco, CA, 94107, United States

Overview

43 is an earliest-stage fund that invests in pre-product startups and works closely with founders to reach product and market fit.

Total investments
5
Lead investments
0
Investments · 12mo
2
Active investors
1
Visit website

Investment portfolio

  • Insight Health

    Participated · Series A · Apr 2026

    Insight Health develops voice-first and chat AI clinical agents that engage patients across touchpoints to perform routine clinical and non-clinical tasks, including phone and front-desk coordination, referral and fax processing, pre-clinical intake, triage, and real-time EHR scribing. The platform is deployed by healthcare clinics such as The Oregon Clinic, Pacific Sports & Spine, Inland Neurosurgery, Coastal Health, and Santiam Hospital and has completed more than 3 million autonomous clinical conversations. The company claims its solutions save clinics hundreds of thousands of administrative hours each month and have contributed to a collective annual administrative cost reduction of over $50 million. Insight Health positions its agents as operating across patient engagement, documentation, and post-procedure triage to improve clinician time and workflow efficiency. Following its $11 million Series A, the company plans to accelerate product development and broaden partnerships with healthcare organizations nationwide.

  • Onton

    Participated · Seed · Dec 2025

    Onton is a San Francisco-based developer of a product search and discovery engine designed to let users look for items with natural-language, image, or combined queries. The platform consolidates product information from multiple websites into a single, streamlined list, aiming to simplify comparison shopping. According to the company, its product already achieves a conversion rate three times higher than the industry benchmark, and more than 20% of its users are weekly active. The business initially focuses on home décor and furniture but plans to expand into apparel and electronics in response to user demand. Recent funding will be used to enhance the product, grow the team, and strengthen the firm’s global reach. To date, Onton has raised approximately $10 million in external capital, including the newly announced seed round.

  • Hypernatural

    Participated · Seed · Jul 2025

    Hypernatural is an AI-powered video creation platform that generates high-quality, bespoke videos in minutes. Its core product offers text-to-finished-video, extensive editing with style updates and regeneration, persistent custom characters, thousands of templates or custom styles, a nearly 2 million-video AI-generated content library, and AI narration and captions. The company serves over a million users and operates a credits-based subscription starting at $12 per month, with higher-tier plans unlocking advanced features. Hypernatural positions itself as a bridge between slow, resource-intensive traditional video production and fragmented AI video models, emphasizing speed, affordability, and consistency. Leadership aims to build a Canva-for-video, with a five-year ambition for AI to be the default way people craft visual narratives. Investors and partners have praised the team’s rapid product development, user growth, and market understanding.

  • Riza

    Participated · Equity · Apr 2025

    Riza builds an "AI-first" infrastructure that enables developers and LLM-powered agents to run code safely and efficiently using a sandboxed WebAssembly (WASM) runtime. The platform supports multiple languages (including Python and JavaScript) and isolates untrusted or dynamically generated code to protect the host environment. Riza emerged from a prototype created by founders Andrew Benton and Kyle Gray (formerly of Twilio, Stripe, and Retool) and has positioned its product for development, CI, and production use. The company has announced general availability and says customers generated over 850 million code-execution requests in March. Riza plans to use its new funding to expand its team, continue work on untrusted code execution, and build more tools to make AI code generation more reliable. The product is marketed around a "Just-in-Time Programming" pattern enabling LLMs to write and execute code in production while reducing manual review and infrastructure overhead.

Team

  • Soso Sazesh

    Co-Founder and General Partner

    LinkedIn