Partner at Kleiner Perkins
Check size: No verified personal check range or current firm-wide standard range is publicly stated on Kleiner Perkins' site. Aditya's official portfolio includes both early and growth investments; round size and signing authority should be confirmed directly.
Aditya backs the infrastructure and applications that make AI useful inside enterprises. His disclosed portfolio spans long-horizon agent infrastructure, AI-first site reliability, conversational data analysis, enterprise accounting automation, generative-internet visibility, model infrastructure, enterprise search, and streaming data. His product and engineering background supports a practical lens on whether a technical capability maps to an urgent workflow and can scale across demanding organizations.
Show the exact enterprise workflow, current failure or labor cost, and why the AI system can own a meaningful outcome. Bring architecture, inference cost, latency, reliability, security, data access, integration, evaluation, and human-escalation design. Demonstrate why usage creates a durable data or distribution advantage and which budget the product replaces or expands.
Technically exceptional teams, infrastructure that makes advanced models reliable and economical, applications that complete meaningful enterprise workflows, a clear data or systems advantage, and products that can become a trusted layer rather than a disposable feature.
No personal pass list is published. His portfolio implies caution toward thin AI wrappers, demos without reliability or workflow ownership, infrastructure without differentiated performance or economics, enterprise tools lacking trust and integration, and security products unable to prove measurable risk reduction.
Ask whether the AI merely suggests an answer or reliably completes an auditable enterprise task, including permissions, integrations, exceptions, and human escalation.
Evaluate model serving, sandboxing, tool access, memory, observability, security, latency, cost, and recovery together for long-running agents.
Determine whether performance, proprietary data, workflow integration, switching cost, or accumulated operational context compounds with usage rather than being erased by the next base-model release.