About Us

We make
sound into insight

Sound into data, data into judgment. Deeply is the eyes and ears of the production floor

Deeply is a deep-tech company that automates with sound AI the judgments people used to make by ear, built on acoustic signal processing and deep learning

  • Deeply presenting its solutions to visitors at Automation World 2026
  • The Deeply team celebrating its anniversary at the new office
  • A Deeply team member inspecting equipment with on-site personnel
  • CEO Suji Lee presenting Machine Hearing at AI EXPO KOREA 2026
  • The new Deeply office entrance with its logo and brand colors
  • Deeply demonstrating Listen AI Safety at SECON 2026
  • Celebrating Deeply’s anniversary with a cake at the office
  • Visitors exploring the Deeply demo at Automation World 2026
  • A Deeply team member working on sound inspection at a manufacturing site

What We Do

DEEPLY Sound AI technology that's changing the world

Technical Strength

Since 2017, focused on nonverbal sound alone

Not speech recognition. Pure audio AI that understands the sounds which are not speech, machine noise and screams and impacts and anomalies, as they are

Why sound

Sound carries what cannot be written down: which machine, under what conditions, and what had gone wrong. Reading that is the work

  1. 01

    An AI that hears better than we do

    Sound is complicated

    • Pattern, frequency and physical signal are mixed together, which makes it harder to work with than vision
    • Not speech turned into text, but the meaning the sound itself carries
    • Thousands of signal-processing experiments and field trials, with researchers ranked among the field leaders by h-index
  2. 02

    Past the problems only sound has

    Unlike vision AI, sound AI has no standard model

    • Noise, echo, real-time response and accuracy all have to be solved at once
    • A decade of R&D got past that
    • Deeply's papers are accepted at ICASSP, the foremost signal-processing conference
  3. 03

    LLMs run on capital, sound AI on groundwork

    Sound AI is built for the floor, and by it

    • Where LLMs compete on GPU capital, sound AI turns on securing closed data
    • It has to know which machine made the sound, under what conditions, and what had gone wrong
    • Not hearing more, but gathering the rare sounds. Accumulated time on the floor is the capability

Customers & Partners

Proven across industrial sites

Industrial customers

  • Hyosung Electric
  • KORAIL
  • Republic of Korea Army
  • Thailand Automotive Institute
  • Hyosung Electric
  • KORAIL
  • Republic of Korea Army
  • Thailand Automotive Institute

Safety customers

  • Sejong Government Complex Gymnasium
  • Incheon Transit Corporation
  • Naejangsan National Park Office
  • IFEZ
  • High1 Resort
  • Lotte E&C
  • Yangcheon-gu Facilities Management Corporation
  • Dobong-gu
  • Seo-gu, Daegu
  • Seongju-gun
  • Sejong Government Complex Gymnasium
  • Incheon Transit Corporation
  • Naejangsan National Park Office
  • IFEZ
  • High1 Resort
  • Lotte E&C
  • Yangcheon-gu Facilities Management Corporation
  • Dobong-gu
  • Seo-gu, Daegu
  • Seongju-gun

Past clients and partners

  • Meta
  • Apple
  • Google
  • Sony
  • Naver
  • Hanwha
  • NII
  • Cappella
  • Korea University Medicine
  • Seoul National University
  • Yonsei University
  • Gwangju Metropolitan City
  • Kangwon Land
  • KOICA
  • LG CNS
  • KB
  • KT
  • Meta
  • Apple
  • Google
  • Sony
  • Naver
  • Hanwha
  • NII
  • Cappella
  • Korea University Medicine
  • Seoul National University
  • Yonsei University
  • Gwangju Metropolitan City
  • Kangwon Land
  • KOICA
  • LG CNS
  • KB
  • KT

Track Record

Where the sound AI has been put to work

  1. 2026

    1. 02

      10+ local governments

      Safety monitoring solution

  2. 2025

    1. 09

      Naejangsan National Park

      Safety monitoring solution

    2. 05

      Hyosung Electric

      Motor drive anomaly analysis

    3. 03

      Company H

      Fastening anomaly sound detection and analysis

  3. 2024

    1. 09

      KORAIL

      High-speed rail machinery anomaly detection

    2. 06

      Incheon Transit Corporation

      Chemical hazard detection

    3. 04

      Sejong Government Complex

      Hazard detection for cultural and sports facilities

  4. 2023

    1. 07

      DL E&C / Lotte E&C

      Construction site noise and hazard monitoring

    2. 05

      Incheon Startup Park

      Hazard detection monitoring service

    3. 04

      Kangwon Land

      Hazard monitoring for High1 Resort

  5. 2022

    1. 07

      ROK Army Headquarters

      Drone engine anomaly sound detection

    2. 05

      KB, ITP

      Fraud prevention and theft detection sound project

    3. 03

      KT

      Elderly care solution, IoT communication project

  6. 2021

    1. 10

      LG CNS

      Speaker emotion detection solution

    2. 10

      LG CNS

      Defence technology collaboration

  7. 2020

    1. 09

      Amazon

      AWS disaster response collaboration

  8. 2019

    1. 06

      Google

      Google for Startups joint solution development

    2. 05

      LG CNS

      AI sound development project

  9. 2018

    1. 09

      Meta

      Asia Startup Program (with Asia Foundation)

Get to us

DEFPLY Head Office

34, Mapo-daero, Mapo-gu, Seoul, 04174, Rep. of KOREA

T.

Adding meaning
to sound
Copyright 2024.. All rights reserved.
Privacy Policy
Terms of Service
Cookie Policy