Experience AI Ops MIHU
AI Ops
It is a chatbot configuration platform designed for Ops teams across various operational products.
What Makes It Unique?
-
Centralized orchestration: Instead of each team building chatbots individually, AI Ops acts as an orchestrator and provides a unified platform.
-
No-code simplicity: Teams can build, test, and deploy chatbots in three simple steps.
-
Reduced engineering load: Ops teams no longer need to request engineering support for chatbot creation.
-
Launch chatbots live without engineering dependency
Impact metrics
The Ask
The AIOps Playground is an experiment environment which enables teams to explore the capabilities of the platform and quickly get an idea of how the platform can help them.
It is meant to be a place where users can easily experiment, test and iterate their proof-of-concept before graduating to the NonProd setup.
The Users
The AI Ops and chatbot are used by tech teams from different Ops domains.
They onboard and test their automation and chatbot use cases here—making operations smoother and more efficient in their areas.”
Current workflow
When a ticket is raised by the platform user regarding any issue...
Design vs Tech direction
Design approach
Intent driven
Automation
AI agents
Proposed workflow
Design direction
Tech direction
User journey
Design direction
Design solution phase I: Tech approach
It aligns with the current technical constraints.
Chatbot: the final touchpoint
A chatbot interface is the final touchpoint for the end-user’s gateway to quick, personalized assistance. Built from configurations in the sandbox, it delivers real-time responses to queries and incident reports, tailored to their respective platform’s data.
Design solution phase II: Design approach
It envisions future possibilities of automation using AI
Automated sanbox experience: Smarter, faster query building
To streamline the chatbot configuration process, the sandbox now features two intelligent agents:
-
Promta: AI agent that automatically generates relevant queries based on the user’s uploaded data, eliminating the need for manual query writing.
-
Optima: AI agent that produces multiple output variations for each query and presents the top three options. Users can select, edit, or finalize the most suitable result.
This two-step automation reduces manual effort, speeds up configuration, and empowers users to choose the best possible outcome without trial-and-error.
As the platform evolved beyond incident resolution into a multi-agent, multi-service ecosystem, the name AI Ops was rebranded to MIHU (May I Help You). This shift reflects a broader mission, offering intelligent, conversational support across diverse operational needs, powered by a network of specialized agents.
A centralized entry point for users to explore the AI Ops platform. The page enables discovery of key services like the query playground and chatbot, facilitates onboarding, encourages contribution, and supports upskilling, all designed to make the platform accessible and engaging for diverse teams.
My journey with AI Ops
I stepped into the AI Ops project as the sole designer, with no PM and only a rough sketch of what the product could be. It was an engineering-led initiative, and collaboration was key, I worked closely with developers, asked the right questions, and shaped the user flow from the ground up.
When technical constraints blocked my original vision, I proposed a phased approach, so we could build what was feasible now, and pave the way for automation when the tech aligned. Together, we built a platform that empowers Ops teams to create, test, and launch chatbots with clarity and confidence.
My understanding
Product teams often struggle to understand how an AI chatbot platform can be tailored to their specific use cases.
Without a safe and flexible environment to simulate user interactions, teams face challenges in experimenting, iterating, and refining their chatbot experiences. This slows down innovation and increases the time to value.
The tech team:
-
Manually analyzes the issue described in the ticket
-
Refers to standard operating procedures (SOPs) to determine the appropriate steps
-
Executes each action step-by-step based on the SOP guidance
Tech contraints
Data insufficiency
No relevant AI model
Time restrictions
Tech direction
.png)
Query sandbox
An interactive sandbox environment where teams can configure, customize and test chatbots tailored to their uploaded datasets.
Highlights:
-
Smart & Contextual: Answers based on end-user's platform’s unique data.
-
Conversational UI: Easy, intuitive interactions.
-
Fast Resolution: Handles common issues instantly, escalates when needed.
A seamless support experience, no tickets, no wait times, just answers.
Extending the Experience:
From Configuration to Identity
AI Ops logo
A visual identity that captures the essence of AI-driven incident resolution. The logo blends Walmart’s branding with symbolic elements of automation and intelligence, reinforcing the platform’s purpose and presence across tools, communications, and experiences.
From AI Ops to MIHU: Scaling with Purpose
Intrasite landing page
