Engineering Director of AI Reliability

I make Qira a reliable and trusted everyday AI companion for millions of users. Over eighteen years, I have designed the architecture of reliable systems—from co-founding startups and delivering multi-cloud software at OX Security and Binah.ai, through global platform operations at Sony and Red Hat, to establishing the global AI reliability organization for Lenovo's Qira ecosystem.

Inbar Rose

About Me

Director of AI Reliability, Lenovo Qira · Unit 8200 Alumnus · 2× Founding Engineer

Inbar Rose

Inbar Rose

Director of AI Reliability, Lenovo Qira

I make Qira a reliable and trusted everyday AI companion for millions of users.

For eighteen years, my career has been defined by a single pursuit: designing the architecture of reliable systems and building the organizations required to sustain them. The trajectory of my work spans the entire spectrum of modern engineering, from the lean urgency of co-founding startups and delivering multi-cloud software from its earliest inception, to orchestrating global platform operations for millions of active users at enterprises such as Sony and Red Hat. Today, I bring this entire background to bear as I establish the global AI reliability organization for Lenovo’s new AI ecosystem, Qira.

My philosophy rests on the principle that reliability cannot merely be patched onto a finished system; instead, we must treat the entire software development lifecycle, from the first draft of a feature specification to its ultimate deployment, as an observable product in its own right. To achieve this, I apply the rigorous discipline of site reliability engineering far beyond the codebase. I connect technical execution with the wider organization, aligning engineering teams with critical partners such as those in legal, procurement, and product. Ultimately, my mandate is to resolve the tension between speed and safety, empowering developers to innovate at their absolute limits while maintaining an unyielding foundation of systemic trust.

EurekaMind Blog

Insights on leadership, productivity, and systems thinking

EurekaMind Substack
Featured Blog

EurekaMind on Substack

Productivity, Growth, Engineering, Learning, Empowerment, Leadership

I write about the intersection of technology, leadership, and productivity. My articles draw from my experience building and scaling engineering teams and products across startups and enterprises.

Latest Articles

Why Truth and Trust Are No Longer the Same Thing
Why Truth and Trust Are No Longer the Same Thing
Jun 8, 2026

Expanding the Survival Envelope
Expanding the Survival Envelope
Feb 20, 2026

Renaissance of Innovation
Renaissance of Innovation
Feb 18, 2026

Good Habits of Agent Orchestration
Good Habits of Agent Orchestration
Jan 26, 2026

The Anatomy of a Mature Agent: A Field Guide
The Anatomy of a Mature Agent: A Field Guide
Jan 9, 2026

Accountability Is a Feature
Accountability Is a Feature
Jan 7, 2026

Predictability is the Point
Predictability is the Point
Jan 6, 2026

Smart Agents, Broken Systems
Smart Agents, Broken Systems
Jan 5, 2026

Escalation Is an Agent Feature, Not a Failure
Escalation Is an Agent Feature, Not a Failure
Jan 4, 2026

What an Agent Can’t Do Matters More Than What It Can
What an Agent Can’t Do Matters More Than What It Can
Jan 4, 2026

Projects

Open source contributions, technical projects, and patents

Agentic Systems Habits
Open Source
AI/ML
Governance
Design Framework

The 7 Habits of Effective Agentic Systems

A design framework that captures seven practical habits that make AI agents safe, useful, and auditable in production. Effective agentic systems are governed components with clearly bounded roles that are embedded into workflows, reducing ambiguity, improving decisions, and providing measurable outcomes.

This framework provides a shared language for engineers, architects, and executive leaders to reason about agentic systems. Rather than optimizing for autonomy, it optimizes for reliability, governance, and system outcomes. The habits are patterns that emerge when agentic systems are designed intentionally and operated at scale.

SPOT Framework
Open Source
Decision Making
Engineering

SPOT Framework

SPOT (Survey, Prioritize, Optimize, Take Action) is a rapid decision-making tool inspired by medical triage. It helps engineers quickly assess, prioritize, and act on tasks, ensuring focus on what matters most in critical situations.

Unlike traditional frameworks like RICE, SPOT is designed for speed, allowing you to make decisions fast without the need for extensive calculations or collaboration. Whether you're handling incidents or managing high-pressure tasks, SPOT helps you take action when every second counts.

Patents

Awarded and pending patents in software engineering and artificial intelligence.

Awarded
Patent
Version Control
Developer Tools

Related Change Analysis of Multiple Version Control Systems

US-20230029262-A1 · Awarded Jan 26, 2023

Select a change in one project and see related changes in other projects.

Pending
Patent
Gaming
AI

Autonomous Video Game Command Execution

Application No. 19/259,275 · Pending

An AI framework allowing non-player characters to autonomously and persistently carry out complex user commands even when outside the immediate view.

Pending
Patent
Gaming
Cloud

Virtual Gaming Session Recovery

Application No. 19/259,216 · Pending

A host-side method to temporarily preserve cloud gaming resources during network drops, allowing players to reconnect without losing session state.

Connect

Reach out via email at inbar@inbarrose.com or connect on LinkedIn.