Durable Sessions AI UX vs IDEO Design Thinking: Which?
// TL;DR
Use the Christensen Durable Sessions Framework if you are building or fixing a real-time AI chat or agent product and need resilient streaming, multi-device continuity, or live agent control. Use IDEO Design Thinking if you face a broad, ill-defined human-centered problem and need to discover what to build before you build it. These frameworks solve fundamentally different problems — one is an engineering architecture pattern for AI delivery infrastructure, the other is a general-purpose innovation methodology. Most AI product teams will need both at different stages.
// HOW DO THEY COMPARE?
| Dimension | Christensen Durable Sessions AI UX Framework | IDEO Design Thinking 7-Phase Process |
|---|---|---|
| Best for | Diagnosing and fixing broken AI chat/agent streaming architectures | Discovering and framing ill-defined human-centered problems |
| Problem stage | You already know what to build but delivery breaks under real conditions | You don't yet know what the right problem or solution is |
| Complexity | High — requires deep understanding of streaming protocols, pub/sub, WebSockets | Moderate — structured but accessible to non-technical teams |
| Time to apply | Days to weeks for architecture redesign and implementation | Weeks to months across all seven phases with iteration loops |
| Prerequisites | Existing AI streaming product, knowledge of SSE/WebSockets, agent architecture | Access to end users for empathy research; no technical prerequisites |
| Output type | Architectural blueprint: session layer design, transport choice, agent topology | Validated solution concept with prototype, insights, and stakeholder narrative |
| Domain specificity | Highly specific to AI/LLM product delivery infrastructure | Domain-agnostic — works for healthcare, software, services, any industry |
| Creator background | Mike Christensen (Ably) — real-time infrastructure engineering | IDEO — global design and innovation consultancy |
| Team composition | Engineering-led: backend, infrastructure, and frontend engineers | Cross-functional: designers, researchers, PMs, engineers, stakeholders |
| Iteration model | Linear audit → redesign → validate cycle | Fluid, non-linear loops across seven phases |
What does the Christensen Durable Sessions AI UX Framework do?
The Christensen Durable Sessions Framework diagnoses why AI chat and agent experiences break under real-world conditions — network drops, multi-device usage, concurrent agents — and provides an architectural solution. It identifies the Single-Connection Trap as the root cause: most AI products stream LLM responses over a direct HTTP/SSE connection tied to one client. When that connection drops, the stream dies.
The framework prescribes a Durable Sessions layer — a persistent, shared, independently addressable resource sitting between agents and clients. Agents write events to the session; clients subscribe to it. This decoupling unlocks three foundational capabilities: Resilient Delivery (streams survive disconnections), Continuity Across Surfaces (sessions follow users across tabs and devices), and Live Control (clients can steer or cancel agent work mid-generation).
It also solves the SSE Resume-Cancel Conflict (closing an SSE connection is ambiguous between disconnect and cancel) and the Orchestrator Dual-Purpose Problem (orchestrators forced to relay sub-agent updates). The natural implementation substrate is a pub/sub channel model with bidirectional transport like WebSockets.
What does the IDEO Design Thinking 7-Phase Process do?
IDEO's Design Thinking process is a general-purpose innovation methodology for solving complex, ill-defined human problems. It moves through seven phases: Frame a How Might We Question, Gather Inspiration through empathy research, Synthesize findings into Insight Statements, Generate Ideas using IDEO's seven brainstorm rules, Prototype at low fidelity, Test against four innovation lenses (Desirability, Feasibility, Viability, Responsibility), and Share the Story to inspire action.
The process is fundamentally about discovering what people actually need before committing to a solution. It emphasizes empathy over assumption, divergent-then-convergent thinking, and creative collaboration from diverse perspectives. It is domain-agnostic and has been applied to healthcare, education, employee onboarding, product design, and virtually every other field where human needs drive innovation.
How do they compare?
These two frameworks operate at entirely different altitudes and solve different categories of problems. They are not substitutes for each other.
The Durable Sessions Framework is an engineering architecture pattern. It assumes you already have an AI product and already know what experience you want to deliver — the problem is that your infrastructure cannot reliably deliver it. It is deeply technical, focused on streaming protocols, transport layers, pub/sub topology, and agent-client decoupling. Its output is an architectural blueprint.
IDEO Design Thinking is a discovery and innovation methodology. It assumes you do not yet know the right problem or the right solution. It is human-centered, research-driven, and collaborative. Its output is a validated concept backed by empathy insights, tested prototypes, and a compelling narrative.
Where they intersect is in AI product development: a team might use Design Thinking to discover that users need seamless cross-device AI conversations, and then use the Durable Sessions Framework to architect the infrastructure that delivers that experience. In this sense, IDEO Design Thinking answers what should we build? while the Durable Sessions Framework answers how do we make it work reliably?
The Durable Sessions Framework is clearly better for teams facing specific AI streaming failures — disconnects, SSE limitations, multi-agent bottlenecks. IDEO Design Thinking is clearly better when the problem itself is unclear, when you need to build empathy with users, or when you need cross-functional alignment on what to pursue.
Which should you choose?
If your AI product already works in demo conditions but breaks in production — users lose responses on mobile, your stop button is unreliable, multi-device sessions are blind to each other, or your orchestrator is drowning in relay logic — use the Durable Sessions Framework. It directly addresses these failure modes with a concrete architectural pattern.
If you are at the beginning of a product initiative and need to understand what users actually need, or if your challenge is broad and ill-defined (e.g., "how do we improve our customer support experience?"), use IDEO Design Thinking. It will help you frame the right question, build empathy, and arrive at a validated concept before you invest in engineering.
Many AI product teams need both, sequentially. Start with Design Thinking to define the right experience, then apply the Durable Sessions Framework to build the infrastructure that can deliver it under real-world conditions. The Durable Sessions Framework solves a problem that Design Thinking would never surface on its own — it is a specialized engineering concern. And Design Thinking solves a problem that the Durable Sessions Framework explicitly ignores — whether you are building the right thing in the first place.
If you must pick one and you are an engineering team with a live AI product that is breaking, choose Durable Sessions. If you are a product or design team exploring a new opportunity space, choose IDEO Design Thinking. There is no false equivalence here: they serve different purposes at different stages.
// FREQUENTLY ASKED QUESTIONS
Can I use Durable Sessions and IDEO Design Thinking together?
Yes, and most mature AI product teams should. Use IDEO Design Thinking first to discover what experience users need, then apply the Durable Sessions Framework to architect the streaming infrastructure that delivers that experience reliably across devices and network conditions. They address different stages of product development.
Do I need to be a developer to use the Durable Sessions Framework?
Yes. The Durable Sessions Framework requires deep understanding of streaming protocols (SSE, WebSockets), pub/sub architectures, and agent infrastructure. It is designed for backend and infrastructure engineers building AI products. Non-technical team members should pair with engineering to apply it.
Is IDEO Design Thinking useful for AI products specifically?
Absolutely. Design Thinking is domain-agnostic and works well for AI products, especially during early discovery when you need to understand user needs around AI interactions. However, it will not address AI-specific infrastructure challenges like streaming resilience or multi-agent coordination — you need a specialized framework like Durable Sessions for that.
What is the biggest mistake teams make when building AI chat experiences?
According to the Durable Sessions Framework, the biggest mistake is coupling stream health to a single client connection — the Single-Connection Trap. According to IDEO, the biggest mistake is jumping to solutions before understanding the real problem. Both are common and costly, but they occur at different stages of development.
How long does it take to implement Durable Sessions?
Expect days to weeks depending on your existing architecture. The core change — inserting a persistent session layer between agents and clients, redirecting agent output to the session, and switching clients to session subscriptions — is a significant infrastructure refactor but follows a clear 10-step workflow.
Can IDEO Design Thinking replace user story mapping or jobs-to-be-done?
Design Thinking complements rather than replaces these methods. Its empathy and synthesis phases can feed directly into jobs-to-be-done frameworks or user story maps. It is broader in scope — covering problem framing through prototyping and storytelling — while JTBD and story mapping are more focused tools within the solution definition phase.
Does the Durable Sessions Framework require WebSockets?
Not always, but it strongly recommends bidirectional transport like WebSockets if you need Live Control (stop buttons, steering messages). SSE alone creates the Resume-Cancel Conflict where disconnect and cancel are ambiguous. For resilient delivery and cross-surface continuity alone, a pub/sub channel model with resumable subscriptions may suffice.
Which framework should a startup building its first AI product use?
Start with IDEO Design Thinking to validate that you are solving the right problem for real users. Once you have a working prototype and begin scaling to production users, apply the Durable Sessions Framework to ensure your streaming architecture survives real-world conditions. Skipping either step leads to building the wrong thing or building the right thing poorly.