What could users try when AI infrastructure startups announced funding?
At pre-seed, Visko let visitors steer a live video and OliverAI offered enterprises a way to test their own queries. At seed, Entire released an open-source CLI that saved coding-agent sessions alongside Git commits.
3 reviewed rounds. Each company described what someone could try on the day it announced funding. None of the three announcements reported how many outside teams used the product in production.
What could someone try on launch day?
Visko · Steer a live Orbis video. A visitor could go to Visko’s website and change the prompt while Orbis was generating a scene. The video kept streaming as it responded to the new direction. Visko left pricing and sign-in requirements unspecified. Its video-quality study did not count returning users. Visko’s launch announcement ·
OliverAI · Test an enterprise workload. A team could bring its own data and queries to OliverDB, either as a managed service or inside the customer’s virtual private cloud. It could compare query speed and compute use on its own workload. OliverAI’s published benchmarks came from its internal tests. The announcement gave no count of customer deployments. OliverAI’s launch announcement ·
Entire · Trace a commit to its coding session. A developer could install the Entire CLI, enable it in a Git repository and commit a change made with Claude Code or Gemini CLI. The resulting Checkpoint linked that commit to its prompts and transcript. The founder listed Codex and Cursor CLI as future integrations and gave no count of teams using Checkpoints. The Entire founder’s launch notes ·