Tale Case Study

Enterprise AI | 2024 - Present | Product Designer

Overview

Tale is a privacy-first platform that enables organizations to run AI agents on their own infrastructure. Instead of relying on external AI providers, companies can automate work, manage customer conversations, build workflows, and collaborate with AI while keeping sensitive data within their own environment.

As the platform evolved, the challenge shifted from designing individual features to designing a system that felt coherent. AI chat, workflow automation, governance, shared knowledge, and customer conversations all needed to work together without overwhelming users or compromising trust.

As Product Designer, I worked across the platform to design new product experiences, evolve the design system, and contribute directly to the production frontend. My focus was making enterprise AI feel understandable, predictable, and practical rather than novel.

Constraints

Most enterprise customers weren't adopting Tale because they were excited about AI. They were looking for ways to automate repetitive work without giving up visibility or control. Many had already experimented with AI products that behaved unpredictably, acted too aggressively, or required sensitive company data to leave their infrastructure.

That changed how we approached the product. Rather than making AI the centerpiece of every experience, we focused on making it feel like a natural extension of existing workflows. Automation was introduced gradually, users always understood what the system was doing, and important actions remained observable before they happened.

That principle shaped decisions across the platform. Workflows could begin through conversation but remained editable before execution, AI-assisted customer replies always stayed under human control, and shared knowledge was designed to grow intentionally rather than automatically. The goal wasn't to persuade users to trust AI. It was to build experiences that earned trust over time.

System view

As Tale expanded, the challenge wasn't adding new capabilities. It was ensuring the product continued to feel like a single system.

Each part of the platform solved a different problem, but users experienced them as one product. That meant consistency became just as important as functionality. Navigation, hierarchy, onboarding, empty states, and interaction patterns all needed to reinforce the same mental model regardless of where users entered the platform.

One example was workflow creation. Early concepts separated conversation from configuration, requiring users to describe what they wanted before rebuilding the same intent manually. We redesigned that relationship by allowing workflows to emerge alongside the conversation itself, making automation visible as it was generated instead of after the fact. Users could review, edit, and confirm every step before anything executed, creating a clearer relationship between human intent and machine action.

Across the rest of the platform, we followed the same philosophy. Rather than introducing entirely new AI interaction patterns, we built on conventions enterprise users already understood, allowing powerful capabilities to feel familiar from the first interaction.

Impact

Working on Tale reinforced something I continue to carry into every product I design: complexity doesn't disappear because features are removed. It disappears when systems are organized well. Whether I was designing new workflows or refining small interactions in production, the goal remained the same: help users understand what the product can do, what it will do next, and why they can trust it. As the platform grew, maintaining that clarity became just as important as introducing new capabilities.