🧔♂️ A friendly human may check it before it goes live. More news here
Khosla Ventures joins $35m round for US AI startup Viven
Viven, a US-based AI startup founded by Eightfold co-founders Ashutosh Garg and Varun Kacholia, has raised US$35 million in seed funding to develop digital twins for workplace collaboration.
The round included participation from Khosla Ventures, Foundation Capital, FPV Ventures, and others.
The startup creates a digital replica for each employee using data from emails, Slack, and Google Docs, allowing co-workers to query these digital twins for project information when colleagues are unavailable.
Viven says its technology uses privacy controls to manage data sharing, and users can review their digital twin’s query history.
Its software is already used by enterprise clients including Genpact and Eightfold.
🔗 Source: TechCrunch
🧠 Food for thought
Implications, context, and why it matters.
Viven’s pairwise privacy claims omit security metrics and certifications
- Viven promotes its pairwise context and privacy system as a fix for sharing issues. It gives no security metrics or hallucination rates (frequency of incorrect model outputs). It also omits audit trail specifications (tamper-evident logging details) and compliance certifications.
- Query history visibility may deter misuse. Enterprise AI deployments require encryption, role-based access controls, and Security Information and Event Management (SIEM) integration to meet GDPR, HIPAA, or SOC 2 standards 1.
- Viven has no published case studies with measurable outcomes. Its deployments at Genpact and Eightfold remain unproven for production use.
- LLM security tools such as Vigil and Rebuff detect prompt injection (malicious instructions hidden in data) and block jailbreaks (methods to bypass model safeguards). Risks increase when digital twins pull sensitive email and Slack data across an organization 2.
AI oversight platforms can meet demand for unified digital twin tracking
- As digital twins spread across workplace tools, teams will need centralized audit trails (comprehensive, immutable logs). These logs should track who queries which twins, what data gets accessed, and whether responses contain sensitive information 3.
- Current LLM observability tools such as Langfuse and Helicone track prompts and latencies. They also track costs and model performance. They miss cross-employee data access patterns that digital twins create, which leaves room for governance-focused startups 1.
- Third-party security vendors could add compliance layers that plug into digital twin platforms. They could flag anomalies when unusual query patterns emerge 3. This mirrors DataSunrise (a database security and auditing product). It enhances Elasticsearch (a search and analytics engine) auditing.
- Investors and operators can build “AI data governance” tools that sit between digital twin platforms and enterprise security infrastructure. These tools should ensure transparency without blocking productivity gains.
Stay updated on the go with our mobile app.
Get latest insights with smoother, more personalized experience through TIA mobile app.




