HADD SCIENCE

[Grand prize] HADD Science’s Woochang Jung proves startup workflow innovation with the AI chatbot ‘Omnis’ 🤖

intro. Omnis: Work Conversations Become Documents. An AI-based project to secure work visibility and resolve documentation bottlenecks. HADD SCIENCE | Jeong Woo-chang, Employee 2026 Regional Industry-Customized Workforce Development Project Excellent Case
PART 01. Background and Current Status
PART 01. R&D Startup's 'Key Pain Points'. Documentation Bottleneck. Researchers spend more time organizing reports than on core experiments. An average of 35 minutes spent organizing per task leads to delayed follow-up decisions.. Lack of Work Visibility. Decision-makers cannot track "who is doing what and to what extent" in real-time. Half of weekly meetings are spent on progress inquiries.
PART 01. Setting SMART Goals. Key Performance Indicator (KPI) / Before / Target / Expected Outcome. Documentation time per task / 35 minutes / Within 5 minutes / -85.7% time savings Decision-maker status awareness delay / Average 2 days / Real-time (<5 minutes) / Maximized decision-making speed
PART 01. Setting SMART Goals. Key Performance Indicator (KPI) / Before / Target / Expected Outcome. Frequency of re-asking for same info / 10 times/week / 1 time/week or less / Reduced communication fatigue Monthly missed tasks / 2.5 tasks/month / 0 tasks / Ensured task completion
PART 02. Core Technology and Process
PART 02. What is learned becomes a function. Prompt & Function. Decomposing work conversations into structured data via prompt engineering. Automatic DB recording with Function Calling.. RAG Knowledge Augmentation. Building an Omnis internal knowledge base. Querying and answering internal information in natural language through Retrieval-Augmented Generation (RAG).
PART 02. What is learned becomes a function. Multi-turn Context. Managing work history in Chat Dock thread format, maintaining complex work context through multi-turn conversation management.
PART 02. 20 Days of Compressed Execution. 2 Days Training + 18 Days Application. Immediately applied to practical work after completing the Gyeonggi-do Economic & Science Promotion Agency's 'AI-based No-code Data Analysis' course.
PART 02. 20 Days of Compressed Execution. Completed system optimized for R&D startups through an ultra-fast cycle of Design (3 days) ➡️ Development (5 days) ➡️ Pilot (5 days) ➡️ Operation (5 days).
PART 02. A Chat Line is a Workflow. 01 Chat Input: Start reports and instructions with a single line of everyday work conversation. 02 AI Decomposition: Automatically extract key information like people, emails, and deadlines. 03 Automatic Registration: Instantly reflected in Omnis work DB and canvas. 04 Real-time Visibility: All employees share status via an integrated dashboard.
PART 03. Achievements and Expected Outcomes
PART 03. Key KPI Improvement Performance (Before vs After). Work documentation time (minutes) Before: 35 minutes After: 5 minutes Status awareness delay (hours) Before: 48 hours (2 days) After: 5 minutes
PART 03. Key KPI Improvement Performance (Before vs After). Weekly report writing (minutes) Before: 220 minutes After: 30 minutes. ※ Based on analysis of 1,342 KakaoTalk texts over 1 month and actual operational averages within the company
PART 03. Economic Value and Reliability. 480hours Annual work time saved (60 person-days). 1,200KRW 10,000 Annual office labor cost savings. Strong verification based on internal data Verified with 1,342 actual internal KakaoTalk conversations, not demo data. The system was demonstrated with anonymized personal and sensitive information to ensure security.
PART 03. SaaS Expansion and Nationwide Rollout. One person's information becomes everyone's information.. Scalability Designed with a modular structure, a SaaS model for bio/materials startups in the Gyeonggi region is planned for verification.
PART 03. SaaS Expansion and Nationwide Rollout. Vision Nationwide expansion of Best of CHAMP+ and presentation of a standard model for AI adoption in SMEs.. Contact: Jeong Woo-chang (woochang4862@gmail.com) | HADD SCIENCE
AWARD 🏆. Grand Prize. HADD SCIENCE Jeong Woo-chang. Chat-based work management system 「OMNIS」 created with AI no-code learning: A 30-day record of simultaneously resolving documentation bottlenecks and lack of decision-making visibility
AWARD 🏆. Award Winners 🏆. Excellence Award Hansol Oriontech Co., Ltd. Kim Ji-baek Case of improving product development capabilities in the shipbuilding sector through electronic circuit understanding education
AWARD 🏆. Encouragement Award Wint Co., Ltd. Kim Jong-taek Case of strengthening personal capabilities and creating corporate performance through the Regional Industry-Customized Workforce Development Project education. 2026 Regional Industry-Customized Workforce Development Project Excellent Case Application Contest
AWARD 🏆. Congratulations to all the winners! 🎉. Winners will be individually notified regarding future schedules and award details. We sincerely thank everyone who showed interest and participated in this contest.
Back to list