What Tariq II has been doing
Tasks, last 30 days
14.2k
For 1,131 people
Needed your approval
5,683
Actions only a person may sign off, last 30 days
Stopped by a rule
197
Refused before the agent saw the data, last 30 days
Open incidents
1
Raised and not yet closed
What went wrong, day by day
Last 30 days, every agent you can see
FailedStopped by a rule
Drafted a follow-up email for the account ownerTariq II for Claire Dubois · On a schedule · 1d agoDone
Cost $0.14 · took 2.5s · model usage 45k tokens in, 538 out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Warehouse QueryCreate Ticket
Ran an approved data query and explained the resultTariq II for Layla Al Mazrouei · On a schedule · 1d agoDone
Cost $0.17 · took 4.1s · model usage 47k tokens in, 2k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Create TicketWarehouse Query
Produced the weekly variance commentaryTariq II for Mariam Al Suwaidi · On a schedule · 2d agoDone
Cost $0.22 · took 3.1s · model usage 52k tokens in, 4k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Warehouse Query
Generated the shift handover packTariq II for Daniel Okafor · On a schedule · 4d agoDone
Cost $0.32 · took 2.6s · model usage 86k tokens in, 4k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Create TicketService Desk Tickets
Answered a policy question and cited the source clauseTariq II for Claire Dubois · On a schedule · 4d agoDone
Cost $0.11 · took 601ms · model usage 30k tokens in, 2k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Create TicketWarehouse Query
Summarised a 34-page document into a one-page briefTariq II for Layla Al Mazrouei · On a schedule · 5d agoDone
Cost $0.17 · took 2.4s · model usage 46k tokens in, 2k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Ran an approved data query and explained the resultTariq II for Layla Al Mazrouei · On a schedule · 5d agoDone
Cost $0.18 · took 2.1s · model usage 54k tokens in, 1k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found 6 possible passages, kept the 3 close enough to use
- 03Drafted the answer and attached where each part came from
- 04Checked its own claims back against those passages
Systems and records it touched
Create TicketWarehouse QueryService Desk Tickets
The model it runs on is not on the approved listTariq II for Claire Dubois · On a schedule · 5d agoStopped by a rule
Cost $0.30 · took 2.5s · model usage 82k tokens in, 3k out
Step by step
- 01Read the request and worked out what kind of task it was
- 02Found a document to work from
- 03A rule stopped it before it read the content
- 04Stopped the task and recorded it
Systems and records it touched
Create TicketService Desk Tickets
Stopped by the rule Personal data never leaves: Names, national IDs, bank details and health data are removed from anything an agent sends outside Mubadala.