At KOGO Tech Labs, I led the design of intelligent dashboards and agent-building platforms that empower enterprise users to harness the full potential of AI. From customizing individual agents to orchestrating multi-agent swarms, these systems were built to serve complex B2B workflows with clarity, control, and scalability. My role focused on translating highly technical AI concepts into intuitive interfaces that support real-time interaction, analytics, and enterprise-grade decision-making.
AI Agent Operating System
Designed AI Agent Dashboards for KOGO’s B2B operating system that enables enterprises to build, configure, and manage intelligent agents. Users can customize agents using features like model settings, guardrails, knowledge bases, and workflows, with built-in analytics and conversation logs.
Swarm Cockpit
Swarm is KOGO OS's intelligent enterprise AI agent system that enables users to leverage advanced AI capabilities through natural conversation. This feature transforms how you interact with your organization's data and knowledge base by providing immediate responses to complex queries.
AI Agent Builder
Led the UX for B2B agent builder that allows enterprise users to create tailored AI agents by configuring tools, workflows, voice, and safety protocols. The platform provides an intuitive interface for managing capabilities and scaling intelligent agent systems. Enabled seamless deployment and monitoring with detailed logs, analytics, and swarm collaboration.
Conversation Analytics Dashboard
Designed a centralized dashboard to help enterprise users track and evaluate conversations between users and AI agents. It visualizes metrics like message volume, resolution rates, intent accuracy, sentiment trends, and failure points. This empowers teams to optimize agent workflows, improve user experience, and fine-tune responses using real-time data and performance insights.
Outcome & Impact
Enabled end-to-end agent management in one unified interface, reducing setup time by 60% and increasing enterprise adoption of AI-powered tools across teams
Streamlined the agent creation process for enterprise clients, cutting onboarding time by 50% and significantly improving team autonomy in configuring AI systems.
I worked on designing a comprehensive AI-powered dashboard for the Ministry of Information and Broadcasting to monitor and analyze media trends efficiently. The dashboard features widgets such as Top Languages, Sentiment Analysis, Top Stories, Leaderboards, Reach & Engagement Rates, Summaries, and Topic Distribution.
With interactive visualizations and a user-friendly layout, the platform delivers actionable insights, empowering the ministry to craft informed policies, enhance communication strategies, and monitor public sentiment effectively.
Outcome & Impact
Enabled data-driven media monitoring and rapid trend response
Improved clarity of public sentiment tracking, supporting more informed government policies
Reduced time-to-insight by 60% through intuitive layouts and smart summaries
Strengthened the Ministry’s ability to manage narratives and public outreach in real time
ASIMOV- Research AI Agent
Crafted the end-to-end UX for Asimov, an enterprise-grade AI research tool that transforms natural language queries into structured research outputs. The system presents insights in a clear, scannable card format and supports contextual Q&A through a conversational AI assistant. Includes a resource library for managing downloadable research documents and PDF archives.
Research Page Overview
The Research Page is the central interface where enterprise users initiate and manage their research queries using natural language. I designed this space to support both quick exploration and deep dives into AI-generated insights.
Key UX features include:
A clean, scannable card layout for research outputs
Clear differentiation between research modes (e.g., quick summary vs. in-depth analysis)
Conversational assistant for follow-up questions and clarification
Visual cues and progress indicators during processing
My Role:
I focused on structuring complex AI content into easily digestible formats. I introduced modular research cards, inline summaries, and clear calls to action like “Read More” to maintain usability across varying levels of content depth.
Internal Research Detail Page – Structured AI Insight View
This page is triggered when a user clicks “Read More” on any research card. It offers a deep, focused look into a particular topic area—ideal for enterprise users validating AI output or preparing reports.
Each detailed view includes:
Title and Topic Context
Key Findings and Bullet Summaries
AI-generated Conclusion
Threaded Q&A interaction with the AI assistant
Citations, source links, and credibility indicators
PDF generation and save-to-library features
My Role:
I designed this page to support cognitive flow—minimizing friction while scanning and cross-referencing. The layout balances depth with simplicity, allowing users to stay in control as they navigate dense AI-generated content. Microinteractions and a sticky sidebar were added to assist in skimming, jumping between sections, and interacting with the AI.
Outcome & Impact
Enhanced the speed and clarity of AI-assisted research for knowledge teams, reducing time spent on manual synthesis by 70% and increasing user satisfaction through intuitive card-based layouts.
Other Projects
Dominar - DIY Feature
Integration of DIY (Do-It-Yourself) module into the KOGO ecosystem of Dominar Bike App to perform basic bike maintenance themselves
Design System | UI Design
I worked as a freelance designer for a finance startup, where I developed their design system, including components, variables, auto-layout, and grids.
