Young-Ju Lee Profile photo Data & AI Engineer · Open to new opportunities

Collect the data,
verify with models,
prove it with results.

I'm an engineer who starts by asking where the data came from. My geophysics research led to a first-author SCIE paper, and at the Samsung SW·AI Academy (SSAFY) I'm building the skills to run LLM data pipelines end to end, from collection to service. What I pursue is results you can trust: I check a model's answer another way before relying on it, and I value being able to explain why a result is right over a result that merely looks good.

Geophysical data researchInternational collaborationLLM · RAG engineeringDeep learning modeling
See workResume PDF GitHub
Looking forData Engineer · AI/ML EngineerLocationOpen to any locationAvailableH1 2027
Scroll — the cards stack one by one

Work

Three chapters — graduate research, undergraduate research and bootcamps. Use the tabs inside each card for individual projects.

01 / 03Research2024.03 – 2026.08

M.S. — Separating gravity & magnetic signals with wavenumber correlation to resolve subsurface structure (SPE Journal, first author)

Designed wavenumber-correlation indices (SLFI · DLFI) that separate overlapping gravity and magnetic signals to map salt and basement structures; completed an SCIE paper and thesis through a Calgary research stay and a 3-institution collaboration

  • Designed wavenumber-correlation indices (SLFI · DLFI) splitting overlapping gravity/magnetic signals; r ≥ 0.94 vs. 3 independent models · code released
  • 6-month stay at Univ. of Calgary, collaboration with Khalifa Univ. (UAE) → first-author SPE Journal paper (SCIE, 2026)
  • Best Student Paper (poster), KSEG 2024 · 3 talks · 10+ lab research-meeting talks · 2025 Basic Research Lab workshop talk
M.S. thesis Fig. 11 — SLFI (basement) · DLFI (salt) maps
M.S. thesis Fig. 11 — SLFI (basement) · DLFI (salt) maps

Wavenumber correlation of gravity & magnetic data for salt/basement structure (M.S. · SPE Journal first author)

2024.03 – 2026.08 · Geophysical Computing Lab · Dept. of Environment and Energy · Graduate researcher

Separated overlapping gravity and magnetic signals in the wavenumber domain to resolve salt and basement structures offshore Abu Dhabi

  • SLFI basement structure agrees with three independent models at r ≥ 0.94 (profile correlation 0.95–0.98 after shift alignment)
  • DLFI reproduces the salt distribution and suggests an additional salt uplift west of Ghasha
Full story →SPE Journal paper ↗Code (gravmag-wcf) ↗
02 / 03Research · AI project2022 – 2024.02

Undergraduate — Seismic interpretation of overseas CO2 storage with plate-tectonic history (first author) · spatiotemporal deep learning capstone

Wrote a first-author paper on seismic interpretation of overseas CO2 storage sites in light of plate-tectonic history, and designed a spatiotemporal deep learning model for wildfire spread

  • First-author paper: GPlates plate-tectonic reconstruction + seismic interpretation of 8 large CO2 storage sites in 4 regions (2024)
  • 4D seismic and CCS case studies, Calgary short stay (Utah FORGE gravity) → 1 second-author paper · 3 talks
  • Capstone: designed a U-LSTM predicting wildfire fronts — Accuracy 0.90 · F1 0.85 on an unseen fire
2024.04 — Utah FORGE poster presentation
2024.04 — Utah FORGE poster presentation

Seismic interpretation of large overseas CO2 storage sites considering plate-tectonic history (first-author paper)

2022 – 2024.02 · Geophysical Computing Lab (undergraduate researcher) · Undergraduate researcher · first author

Reconstructed how large overseas CO2 storage sites formed and interpreted their seismic images from a structural-geology perspective

  • First-author paper: Analysis on Seismic Interpretation for Overseas Large-scale CO2 Storage Considering Geological History Related to Plate Tectonics, Geophysics and Geophysical Exploration 27(1), 1–22 (2024)
  • Showed that stratigraphy in each seismic image reflects the region's tectonic history → a basis for rejecting geologically unrealistic velocity models when processing domestic data
Full story →First-author paper ↗
03 / 03AI project2025.06 – present

Bootcamps — SSAFY 16th cohort · POSCO Youth AI & Big Data Academy 30th cohort

LLM data pipelines, a Django service and a VQA model ensemble at SSAFY; a Konglish-correction generator with RAG at the POSCO academy

  • SSAFY AI Challenge (VQA): Qwen/Gemma QLoRA, OCR, multi-view probability ensemble → public 0.96961
  • ThRead: crawling → LLM judgments verified in code → pandas pipeline, Django backend · 86 tests
  • Heritage RAG guide (solo): public-API & Wikipedia collection → LangGraph RAG that shows only citations verified in code · 50 offline tests
  • POSPEAK: correction datasets, FLAN-T5 LoRA strategy comparison (BERTScore F1 0.32 → 0.74) + RAG · award
Heritage RAG guide — pick a national treasure, ask, and see only the evidence actually cited (offline demo mode, no API key)
▶ Demo — Heritage RAG guide

SSAFY — ThRead LLM comment-analysis pipeline · AI Challenge VQA ensemble

2026.07 – present · Samsung Software & AI Academy for Youth (SSAFY) · team of 2 · Backend & data pipeline · model training (AI Challenge) · Team of 2

Extended a reading-community service across five projects, from data collection to LLM analysis and web service, plus a multimodal QA challenge

  • AI Challenge public score 0.94608 (4B baseline) → 0.96961
  • 86 automated tests passing; all project-5 requirements implemented, plus progress streaming and analysis history
Full story →▶ DemoGitHub ↗Heritage RAG guide ↗

Footprints

Where I worked and the regions I studied with data — tap a point to see what I did there.

← Swipe the world map sideways →

Western Canada Abu Dhabi · Persian Gulf Utah FORGE geothermal site North Sea Gorgon

Korea (zoomed)

Jeonbuk National Univ. Uljin–Samcheok wildfire Gunwi wildfire Miryang wildfire Gangneung wildfire

Details

Tap a point on the map.

  • Where I worked · 3
  • Regions I studied with data · 16
  • Partner institution · 1

Locations are approximate; offshore fields and storage sites are shown at a nearby point. Countries outlines: Natural Earth via world-atlas.

Background

Education · publications · awards · skills

Education

Samsung SW & AI Academy (SSAFY)16th cohort Training program · 2026.07 – present
Jeonbuk National Univ. (Graduate)Environment and Energy M.S. · 2024.03 – 2026.08
POSCO Youth AI & Big Data Academy30th cohort Completed · 2025.06 – 2025.08
Jeonbuk National Univ.Mineral Resources and Energy Engineering B.S. · ~ 2024.02

Publications

Wavenumber Correlation of Gravity and Magnetic Anomalies for Constraining Salt/Basement Structure in Offshore Abu DhabiSPE Journal (SCIE) · first author · 2026
Analysis on Seismic Interpretation for Overseas Large-scale CO2 Storage Considering Geological History Related to Plate TectonicsGeophysics and Geophysical Exploration 27(1), 1–22 · first author (in Korean) · 2024
Geological Analysis of the CarbonNet CCS Project in the Gippsland Basin, AustraliaGeophysics and Geophysical Exploration · second author (in Korean) · 2023
conference talks 6

Awards · Certifications

Project Excellence Award, POSCO Youth AI & Big Data Academy (30th)POSCO HRD Center · POSTECH · 2025.08
Best Student Paper (poster), KSEG Fall Conference 2024Korean Society of Earth and Exploration Geophysicists · 2024.11
ADsP (Advanced Data Analytics Semi-Professional)Korea Data Agency · 2022.09

Skills

Click a skill to see where I used it

Languages & frameworks
AI & data analysis
Spatial & geophysical data
Domain

Looking forward to
working together.

iyju990312@naver.com

Send emailFull resumeLinkedIn

No LLM — answers are passages found on this site, each with its source. Nothing is generated.