Cord Messaging Ai
cord is the only job search platform that makes finding work direct, transparent and human.
Role:
Sr. Product Designer
Service Provided
UX Research, Wireframing, UI Design, Prototyping
The problem(s) and my role:
Cord needed a more intelligent, efficient, and context-aware messaging system to improve communication across its platform. As teams scaled and message volume increased, users were struggling with: Information overload – too many messages, difficult to prioritize or respond efficiently. Lack of context – users often received messages without enough background, leading to delays or miscommunication. Fragmented workflows – switching between tools or threads to complete tasks slowed down collaboration. Low responsiveness – important messages often got buried, and follow-ups were inconsistent. The goal was to design an AI-powered messaging system that could streamline communication, reduce noise, and enable faster, smarter interactions across teams. As the sole designer on the project, I led the design process from discovery through to implementation. This included: User Research Conducted interviews and workflow analysis to understand pain points around current messaging practices. Synthesized findings into personas and user journey maps to highlight communication bottlenecks. Product Strategy & UX Design Collaborated closely with product and engineering to define the core functionality of the AI system, including features like smart summaries, predictive replies, intent recognition, and intelligent nudging/reminders. UI Design & Prototyping Designed high-fidelity UI components for AI-assisted interactions, ensuring a seamless, unobtrusive experience that blended naturally into existing workflows. Created interactive prototypes to validate ideas quickly with stakeholders and test with users. AI Behavior Design Worked with machine learning engineers to shape how the AI would behave—tone of voice, confidence levels, when/how to intervene—balancing utility with trust and usability. Iterative Testing & Refinement Ran usability tests across multiple iterations, continuously refining the experience based on feedback. Focused on ensuring the AI added value without overwhelming or distracting users.
1
The Challenge:
🔎 1. Tackling Information Overload
Problem: Users were overwhelmed by high message volume and struggled to identify what needed attention.
Solution:
Introduced AI-powered message prioritization, visually elevating high-importance messages based on urgency, relevance, and user behavior patterns.
Designed smart filters and highlights that allowed users to instantly surface tasks, questions, and blockers in conversation threads.
Integrated summarization cards for long threads, enabling users to get a quick overview without reading every message.
Created a AI settings panel that allows the user to turn off AI assistance if needed.
🧠 2. Bringing Context to Conversations
Problem: Users lacked context when joining threads, leading to miscommunication and time wasted getting up to speed.
Solution:
Designed an AI-generated “catch-up summary” that automatically appeared when users returned to a thread they hadn’t seen in a while.
Included in-line knowledge references, pulling in linked documents, tasks, or past decisions to provide real-time context in conversations.
Used progressive disclosure patterns to surface relevant details without cluttering the UI.
⚙️ 3. Reducing Workflow Friction
Problem: Users had to switch between tools or follow-up reminders manually, breaking their focus.
Solution:
Integrated task creation directly from messages, so users could turn action items into tasks without leaving the thread.
Designed AI nudges for unresolved threads or missed replies, helping teams stay aligned without constant manual follow-up.
✍️ 4. Designing AI that Feels Human
Problem: Users were hesitant to trust or rely on automated suggestions if the AI felt too intrusive or robotic.
Solution:
Worked closely with engineers to define the AI’s tone of voice—confident but polite, helpful but not pushy.
Created an area that allows the user to select what kind of tone they want the message to generate such as, polite, warm etc…
Designed transparent AI behaviors, such as explainability prompts (“Why am I seeing this?”) to build user trust.
Focused on opt-in control, giving users agency over how and when AI features were used—ensuring it augmented, not replaced, their workflow.
Allowed the user to only use AI to refine their message if requested.
2
The Result
Designing Cord’s AI messaging system was a deeply rewarding challenge. As the sole designer, I was responsible for crafting a solution that blended powerful AI capabilities with thoughtful, human-centered design. By focusing on real user pain points—information overload, lack of context, and workflow inefficiencies—I was able to deliver an experience that not only improved productivity, but also built trust and confidence in AI within everyday communication.
The success of this project demonstrates the impact of design-led thinking in AI product development, and reinforces the value of embedding usability, clarity, and empathy into even the most advanced technologies.
3
