D-SKET · Events B2B · MICE

When the event ends,
the insight begins.

Turn the customer data that disappears after an event into living insight. Behavior sequence analysis, anomaly detection, and automatic natural-language reports. A data platform for MICE and B2B marketing.

Request a demo → Explore features
01 · THE PROBLEM

The moment the event ends, the data disappears too

Check-ins, booth visits, session attendance, surveys, payments — tens of thousands of data points are generated during an event, but most of them stop at a spreadsheet. That's why measuring ROI, and improving the next event, is so hard.

77%

of post-event data goes unused in analysis
Source: Harvard Business Review, 2023

76%

of B2B marketers say they can't accurately measure event marketing ROI
Gartner CMO Survey

₩20T

annual size of Korea's MICE industry
Korea Tourism Organization, 2024

02 · CORE FEATURES

5 core features

DASHBOARD · MOCK

SK CTO Summit 2026 · Live

LIVE · 1,247 people
Check-ins
1,247
Booth visits
8,902
Session attendance
87%
Conversions (demo requests)
142
⚠ ANOMALY DETECTED · 14:32

Simultaneous movement detected across Hall B booths 1, 2, and 3 — 13 people on the same route. Recommend classifying as an interest cluster.

Behavior sequence analysis

Behavior Sequence Embedding

Converts attendee movement into vectors to automatically form clusters of similar behavior. Discover "interest categories" live during the event.

Anomaly detection

Anomaly Detection

Real-time alerts for chip-movement and dispersion patterns that deviate from the usual behavior distribution. Spot popular booths and manage crowding at the same time.

NLG reports

Automatic natural-language reports

The morning after the event, a one-page executive summary plus a 30-page detailed marketing report are emailed automatically.

Predictive model

Customer conversion prediction +23%

Estimates the probability of conversion within 6 months from behavior data during the event. It tells your sales team who to reach out to first, starting next week.

03 · TECH STACK

Infrastructure for large-scale events

TIME-SERIES

TimescaleDB

500K records per event, with millisecond queries. PostgreSQL-compatible.

ML

Sequence Embedding

An in-house trained behavior-vector model. Similarity-based clustering.

LLM

GPT-based NLG

Automatic data-to-natural-language report generation. Korean and English output side by side.

04 · IMPACT

The impact of adoption

0%
Client cost savings (labor + agency → subscription + operations)
0%
Higher customer conversion rate (after adopting the AI predictive model)
0%
Faster decision-making (with the real-time dashboard)
05 · GET STARTED

Getting started

D-SKET Events is a custom B2B deployment. Pricing depends on the scale of your event and the scope of integration with your existing systems (CRM, payments, access control).

Start with a 2-week PoC

A free PoC using data from one real event. If you're happy with the results, move to a full contract.

Turn discarded data into an asset.

Start with a free 2-week PoC.

Apply for a PoC →