Blog
Notes from the Vektor team
Engineering posts on agent evaluation, tracing, memory, and the messy reality of shipping AI to production users.
TechnicalJan 28, 2025 · 8 min read
Designing Reliable Multi-Agent Workflows
Failure modes compound across orchestrated agents. The evaluation patterns we use to keep multi-agent systems honest — handoff schemas, loop detection, and memory coherence.
A
Aryan Mehta
CEO, Vektor
ProductJan 22, 2025 · 4 min read
Why We Rebuilt Memory From Scratch
Memory v1 was a thin wrapper over a vector store. Memory v2 is a real system — three first-class stores, selective compression, and optimistic locking across agents.
L
Lena Kovacs
CTO, Vektor