LensReading which lens this session carries.
RD5A

Zoom 3 of 3 — line

Model Development

Transactional line · Research & Development · Data Science & Modeling

Runs when: Runs transactional. Starts when a scheduled window opens.

2326

units a week

76%

touchless

8

stations

16.1h

timetable

13.6h

actual

-16%

vs timetable

3 clear4 evidenced2 held

Station walk

Every stage, who owns it, and what is sitting there

Each station names the agent that runs it and the layer it belongs to. Gates name the rule and the human approver. Trains parked at a station are live units of work — click one to read its full evidence trail.

  1. 01

    Frame problem

    Workflow

    run by Feature Engineering Agent

    dwell 44msla 60mqueued 0
  2. 02

    Engineer features

    AI GBS AI GBS depot

    run by Feature Engineering Agent

    dwell 152msla 271mqueued 31
  3. 03

    Train

    AI GBS AI GBS depot

    run by Feature Engineering Agent

    dwell 89msla 210mqueued 13
  4. 04

    Validate

    AI GBS AI GBS depot

    run by Feature Engineering Agent

    dwell 179msla 237mqueued 27
  5. 05

    Explainability pack

    Workflow

    run by Explainability Agent

    dwell 83msla 147mqueued 4
  6. 06

    Reviewer approve

    Policy

    run by Director, Data Science

    Gate: Held for a named human when the unit exceeds the model decision envelope, when confidence falls below the floor, or when a policy clause is engaged.

    approver Director, Data Science

    dwell 81msla 144mqueued 9
  7. 07

    Register

    AI GBS AI GBS depot

    run by Feature Engineering Agent

    dwell 130msla 181mqueued 2
  8. 08

    Deploy

    Policy

    run by Director, Data Science

    Gate: Held for a named human when the unit exceeds the model decision envelope, when confidence falls below the floor, or when a policy clause is engaged.

    approver Director, Data Science

    dwell 181msla 410mqueued 8

Live trains on this line

9 units in motion right now

RefWorkHeld byValueStatusAge / SLA
RD5A-1005

Model at engineer features

Pierre Bianchi

DATA-ANALYTICS spine

gbs

$59kEvidence posted3h 25m / 4h 31m
RD5A-1009

Model at engineer features

Kofi Ellery

DATA-ANALYTICS spine

gbs

$41kEvidence posted2h 48m / 4h 31m
RD5A-1018

Model at deploy

Pierre Nakamura

Model Training Agent

agent

$61kEvidence posted6h 18m / 6h 50m
RD5A-1024

Model at explainability pack

Amara Sorensen

Validation Agent

agent

$20kIn flight1h 22m / 2h 27m
RD5A-1031

Model at reviewer approve

Yuki Baptiste

Model Training Agent

agent

$54kEvidence posted1h 11m / 2h 24m
RD5A-1038

Model at reviewer approve

Zara Ito

Director, Data Science

human

$59kHuman hold27m / 2h 24m
RD5A-1046

Model at explainability pack

Sofia Lindqvist

Modeling Orchestrator

agent

$31kIn flight1h 24m / 2h 27m
RD5A-1053

Model at validate

Pierre Ellery

Director, Data Science

human

$2kHuman hold6h 4m / 3h 57m
RD5A-1060

Model at validate

Zara Baptiste

DATA-ANALYTICS spine

gbs

$3kIn flight3h 13m / 3h 57m
Back to the Research & Development line map