IT Staff OÜ
IT Staff says it engineers quality into AI-infused systems: making AI-driven and distributed systems testable, reliable, and production-grade. Its site also presents the business as engineering consulting for production AI and related software systems.
At a Glance
- Teams building AI-agent and Model Context Protocol infrastructure
- Java and Kotlin/JVM engineering teams
- Companies operating AI-infused or distributed production systems
- Developers needing framework-independent, serverless, or highly concurrent MCP deployments
AI Tools by IT Staff OÜ
(1)
Tachyon MCP
Java MCP Server Runtime
Discussions
No discussions yet
Be the first to start a discussion about IT Staff OÜ
Latest News
Tachyon 1.0.0-beta.31 released with extension-negotiation, CORS/origin, SSE buffering, HTTP/1.1 pipelining, and task-session fixes.
Tachyon 1.0.0-beta.30 released with session-store optimization, MCP header-mirror support, observability improvements, Spring Boot configuration changes, Kotlin DSL changes, and SkillsExtension packaging.
Tachyon 1.0.0-beta.28 released with expanded Kotlin DSL/example work and ongoing protocol, extension, and runtime updates.
Tachyon 1.0.0-beta.27 released as the project continued its rapid beta cadence and MCP protocol/runtime development.
Products & Services
An open-source Java 21+ and Kotlin Model Context Protocol server runtime built on Netty. It provides Streamable HTTP transport, protocol-version support, virtual-thread handlers, tools, resources, prompts, completions, tasks, elicitation, extensions, sessions, and observability; it is distributed under Apache 2.0 through GitHub and Maven Central.
IT Staff's website positions the company around making AI-driven and distributed systems testable, reliable, and production-grade, with related software-engineering and recruiting/sourcing guidance.
Market Position
IT Staff positions Tachyon as a spec-forward, standalone pure-Java MCP runtime: more than a protocol library but without requiring a framework. Its own comparison names the official MCP Java SDK, Spring AI MCP, and Quarkus MCP as adjacent options; Tachyon differentiates with Netty transport, virtual-thread execution, first-class extensions, Kotlin support, and rapid tracking of MCP drafts and SEPs.
Leadership
Founders
Kira Churakova
Founder, shareholder, and management-board member of IT Staff OÜ. Public company records identify her as a founder; a prior-role biography was not located.
Konstantin Pavlov
Co-owner and engineering lead at IT Staff OÜ; software engineer focused on Java, Kotlin, JVM architecture, and AI. His public profile lists prior roles including Staff Software Engineer at Kerno, Technical Lead at JetBrains, and Senior Software Engineer at Twilio Inc.; he also authors Tachyon MCP.
Executive Team
Kira Churakova
Management Board Member / Founder
Public Estonian company records identify her as a founder, shareholder, and management-board member of IT Staff OÜ.
Konstantin Pavlov
Management Board Member; Co-owner and Engineering Lead
Software engineer focused on JVM architecture and AI; public profiles list Kerno, JetBrains, and Twilio experience and identify him as the author of Tachyon MCP.
Board of Directors
Founding Story
IT Staff OÜ was registered in Estonia on 11 January 2024. The company describes its initial direction as engineering quality into AI-infused systems, with an emphasis on making AI-driven and distributed systems testable, reliable, and production-grade; its public material also covers software-engineer sourcing and engineering consulting.
Business Model
Revenue Model
The public site describes engineering consulting for production AI and software systems; Tachyon is an Apache 2.0 open-source runtime distributed via GitHub and Maven Central. No public paid Tachyon pricing was found.
Target Markets
- Teams building AI-agent and Model Context Protocol infrastructure
- Java and Kotlin/JVM engineering teams
- Companies operating AI-infused or distributed production systems
- Developers needing framework-independent, serverless, or highly concurrent MCP deployments
- Building production MCP servers for AI agents on the JVM
- Java or Kotlin applications that need tools, resources, prompts, completions, tasks, or elicitation
- Framework-independent MCP deployments and serverless/stateless tool-calling workloads
- Long-lived or bursty AI-agent traffic requiring Netty transport, sessions, replay, and backpressure
- Engineering and testing AI-infused and distributed systems