Enigma, a robotics AI startup founded by former cybersecurity researchers, secured $71 million in seed funding led by Index Ventures and Ribbit Capital to develop foundation models that reduce the training data required to deploy robots across different environments. The company's models can be adapted to new robots with significantly less video footage than traditional approaches, addressing a key cost barrier in robotics AI where collecting task demonstrations typically requires expensive mock production setups. This capability enables faster and cheaper robot deployment while Enigma builds out its interface platform and recruits engineering talent to scale its models across diverse robotic hardware.
Enigma, a new robotics startup founded by two former Microsoft and Israeli cybersecurity researchers, raised $70 million to develop intuitive human-robot interfaces by studying how people naturally interact with machines. The company is launching a public experiment allowing anyone online to control over 100 proprietary robots in Israel and California performing tasks like painting and chemistry experiments. Enigma aims to make robot control as intuitive as turning a volume knob, shifting focus from pure model capability to interface design and human-centered interaction patterns.
Roboto Agents can now trace robot failures from logs directly to the responsible code and propose fixes, combining multimodal sensor data analysis with repository access through the MCP protocol. BRINC reduced complex flight failure diagnosis time from hours or days to minutes using the system. This enables faster root cause analysis by consolidating investigation work that previously required bouncing between support, engineering, and repository teams.
Encord, a data tooling company, is experimenting with brain wave sensors and muscle electrical signals to improve physical training data for humanoid robots, working with German startup Zander Labs. The company estimates that densely annotated physical training data is worth 100 times as much as basic video data but costs 20 times more to produce. The scarcity and high cost of real-world manipulation data has become a significant bottleneck for robotics companies developing physical AI models, creating a new business around manufacturing training data that doesn't exist at scale.
Every AI story that matters,
in your inbox by 8am.
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the
day in two minutes. Follow companies and topics for alerts, or get the
briefing in Slack. Free, no spam, unsubscribe anytime.
Reading TLDRocket needs no cookies, and the readership counts we rely on come from
our own cookieless analytics. Google Analytics is the exception: it sets cookies and
reports to Google, so it stays switched off until you allow it. You can change your
mind any time from “Cookie settings” in the footer.
Strictly necessary
Session security and form protection (tldrocket-session,
XSRF-TOKEN, 2 hours). The site cannot work without them,
so they need no consent.
Always on
Google Analytics 4 (_ga,
_ga_<id>, up to 2 years). Measures which
stories and sections readers use. Google acts as a third-party processor and may
store the data outside the EU. No advertising, no profiling, no data sold.
