Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Generative UX in LLM Application
Search
Hoon Heo
September 21, 2023
Technology
1
970
Generative UX in LLM Application
Generative AI 시대에 알아두어야 할 Chat Interface 그리고 사용자 경험 관련 추가 고려 사항
Hoon Heo
September 21, 2023
Tweet
Share
More Decks by Hoon Heo
See All by Hoon Heo
What if...? 처음부터 다시 LLM 어플리케이션을 개발한다면
huffon
0
2.1k
고군분투 LLM 프로덕트 적용기: Blind Prompting 부터 Agent까지
huffon
3
1.8k
Autonomous Agent in Production
huffon
2
1.1k
Other Decks in Technology
See All in Technology
リファクタリングへの耐性が高いモデルベースの統合テストの紹介 / Model-Base Integration Test for Refactoring
yuitosato
5
1.5k
Databricksワークショップ - 生成AIとDWH
taka_aki
2
4.5k
とあるユーザー企業におけるリスクベースで考えるセキュリティ業務のお話し
4su_para
2
280
Capybara+生成AIでどこまで本当に自然言語のテストを書けるか?
yusukeiwaki
6
1.3k
顧客が本当に必要だったもの - パフォーマンス改善編 / Make what is needed
soudai
21
6k
AI Builder について
miyakemito
1
140
Emacs x Nostr
hakkadaikon
1
130
生成AIとAWS CDKで実現! 自社ブログレビューの効率化
ymae
2
200
WINTICKETアプリで実現した高可用性と高速リリースを支えるエコシステム / winticket-eco-system
cyberagentdevelopers
PRO
1
180
小規模に始めるデータメッシュとデータガバナンスの実践
kimujun
3
370
30万人が利用するチャットをFirebase Realtime DatabaseからActionCableへ移行する方法
ryosk7
2
300
カメラを用いた店内計測におけるオプトインの仕組みの実現 / ai-optin-camera
cyberagentdevelopers
PRO
0
100
Featured
See All Featured
Testing 201, or: Great Expectations
jmmastey
38
7k
Practical Tips for Bootstrapping Information Extraction Pipelines
honnibal
PRO
9
670
Done Done
chrislema
181
16k
How to Think Like a Performance Engineer
csswizardry
19
1.1k
ReactJS: Keep Simple. Everything can be a component!
pedronauck
664
120k
Unsuck your backbone
ammeep
668
57k
Facilitating Awesome Meetings
lara
49
6k
Automating Front-end Workflow
addyosmani
1365
200k
Navigating Team Friction
lara
183
14k
What’s in a name? Adding method to the madness
productmarketing
PRO
22
3.1k
RailsConf 2023
tenderlove
29
870
The Myth of the Modular Monolith - Day 2 Keynote - Rails World 2024
eileencodes
13
1.9k
Transcript
(FOFSBUJWF69 JO--."QQMJDBUJPO )PPO)FP!-*/&3
ঢ়զঢ়զী
(15ۄח֙ णפ
None
ܻחӒ৬ۧѱ ࣗా೮যਃ
ݺ۸ਃ ݓۅࠁ
Ӓ۞ݶӒח ۠ߑधਵ۽೮ભ
ݺ۸ਃ ݓۅࠁ ߸
t۠ ߸݃ীٜঋ֎ਃu
tਃޙҳܳ ࣻ೧ঠѷযਃu
ࣻػݺ۸ਃ ݓۅࠁ
ࣻػݺ۸ਃ ݓۅࠁ ࣻػ߸
tѪب݃ীٜঋইਃu
ࣻػݺ۸ਃ ݓۅࠁ
ࣻػݺ۸ਃ ݓۅࠁ ࣻػ߸
tӈଳu
t҅ࣘۧѱ ࠗఠदࠛಞೞѱ ࢎਊ೧ঠೡөਃ u
ӝഥܳನೠ झఋসٜ١
$PQZBJ +BTQFSBJ
ૣও؍ (158SBQQFS ࢿद
ܳѢ
None
࢜۽ੋఠಕझ١
࢜۽ੋఠಕझ١
ݺ۸ਃ ݓۅࠁ
߸
t݃ীٜঋইਃu
୶оݺ۸ਃ ࢎਊਃҊ۰ػ ࢜۽߸
None
դਃӝরغח োযӝ߈ೖ٘ߔ
'BNJMJBS$POWFSTBUJPO 4IPSU'FFECBDLMPPQ *UFSBUJWF.VMUJUVSO
ڪযդӝࣿ۱җࣼೠ69؋࠙ীױӝрীষդࢿҕਸѢك$IBU(15
৬ೣԋࠛӝदೠ $IBU*OUFSGBDFৌೂ
None
None
җো$IBU*OUFSGBDFо (FOFSBUJWF69ࠗੌө
(FOFSBUJWF"*ܳࢎਊೞחઁಿ ݽف$IBU*OUFSGBDFৈঠ݅ೡө
/PQF
Ӓۧݶ
$IBU*OUFSGBDFо צডޖੌө
#MBOL1BHF4ZOESPNF
None
tযڃਃਸܻ೧ࣻ u tয٣өঌҊ u tযڌѱޛযࠊঠೞ u j
None
t$IBU*OUFSGBDFо מࢎחইפҳաu
-FUsTHPEFFQFS
None
ݓۅҗبਃࢿ GSPN-JOVT-FF
эҊ۰ооػ (FOFSBUJWF69ӝ߈ઁಿ
74$PEF$PQJMPU
ࢎਊח
tೣࣻݺuۄחݓۅೞী $POUFYU
Oୡрݠޛݴ٘оࢤࢿغӝܳӝೞҊ $POUFYU *OUFOU
ઁಿࢎਊоਗೞחѾҗܳ߈ജ $POUFYU "DUJPO *OUFOU
য٣өঌҊਵݴ ޖਸӝೡࣻਸ যڌѱ೯ز೧ঠೞחоݺഛೠઁಿ
*OUVJUJWFOFTTJTEPXO
ৈӝөо
աݠח
ࢤпਸ೧ࠇद
ࢎۈٜ$PQJMPUীѱ ٘ࢤࢿ݅ਸਗೡө
ޙࢲചח
ܻಂష݂
٘୭ചח
ਫ਼Ӭ
ݽٚਃҳࢎ೦ਸ ੌੌӝמ୶о۽ ೧ঠೡө
None
'MFYJCJMJUZJTUPHP
*OUVJUJWFOFTTBOE'MFYJCJMJUZ GSPN-JOVT-FF
*OUVJUJWFOFTTBOE'MFYJCJMJUZ GSPN-JOVT-FF
$IBUBTBO&TDBQFIBUDI
None
None
*OUFOUীӡࣻח ݆ࣻਬૉாझ
оמೠਬ
(FOFSBMJ[BUJPOUISPVHI*OTUSVDUJPO5VOJOH
ۢ$IBU*OUFSGBDFܳ ࢎਊೞݶ
ݽٚਃҳࢎ೦ਸ ӝמ୶о۽ ೞঋਸࣻҊ
ܻо࢚ೞޅೠ ୶оੋਊ۹ܳߊѼೡࣻب 4FSFOEJQJUZ
ࢎਊ৬ೣԋೞח&YQMPSBUJPO
ߓ
ݓۅࠁח݆ਸࣻ۾જ
ࢎਊ৬ઁಿ эಕীযঠೠ
ӝמبоݺഛ೧ঠೠ
$IBU*OUFSGBDFܳ ࠺࢚ҳ۽ഝਊ೧ਸ־ܽ
ਊ
#SPXTJOH$PQJMPU ࢎ۹
ݓۅࠁח݆ਸࣻ۾જ
$POUFYU
ࢎਊ৬ઁಿ эಕীযঠೠ
None
ࢎਊоࣗ࠺ೞҊחబஎܳ ઁಿب೧ೡࣻਸ ঌܻӝਤೠѐࢶ
ഝࢿചغਸٸ ഐߡ೮ਸٸ ӝמࢎਊद
ӝמبоݺഛ೧ঠೠ
ഝࢿചߡౡ୭ୡݽण
ഝࢿചߡౡ୭ୡݽण
ӝמ೧بࠗ t߄۽ബਊਸઁҕೡࣻחӝמਵ۽߸҃u
None
None
ݺഛೠ69ઁҕਸా೧ѐࢶ
$IBU*OUFSGBDFܳ ࠺࢚ҳ۽ഝਊ೧ਸ־ܽ
ݺഛೠӝמઁҕਸా೧ *OUVJUJWFOFTTܳഛࠁ೮݅
4VNNBSZ оձ tژܲu݃
tח߄۽ޙਸೞҊरযਃu tݒߣਃডաয়פࠗझ۞ਕਃu
'MFYJCJMJUZJTUPHP
None
None
None
t5PPMUJQীبਊೞu
None
tࢎਊٜ যڃਃਸೞҊ u
None
tBOTXFS u
ੋఠ֔ীࢲऔѱ݅զࣻחޙઁ೯
None
$IBU*OUFSGBDFܳా೧ ߊѼೠ4FSFOEJQJUZ
ܻ
ݽ࢚ٚടীԙٜযݏח হ
$IBU*OUFSGBDFب t߽݅ాডuইפ
(FOFSBUJWF"*दীب दա
بݫੋ ઁಿ١ীঌݏ 69ܳҊೞҊਊ೧աоঠ
t69JOOPWBUJPOJTOFWFSFOEJOHu
&0%
(FOFSBUJWF*OUFSGBDFT#FZPOE$IBU /BUVSBMMBOHVBHFJTUIFMB[ZVTFSJOUFSGBDF *OUFSBDUJOHXJUI--.TXJUI.JOJNBM$IBU 0QFODIBMMFOHFTJO--.SFTFBSDI *OUFSBDUJPO$FOUSJD"* ଵҊ