Original Message:
Sent: 02-27-2026 08:13
From: Balaji Balakrishnan
Subject: Technical Query: Achieving Direct Agent Transfer with ACW using Standard Routing
Hi Phaneendra,
Thanks for your suggestions and workaround solutions.
I appreciate the insight regarding skill-based eligibility, but in this case, we are using non-skill routing. Additionally, the agents are not static; they are dynamically configured at a Data Table level, so we aren't targeting just one fixed individual across the board.
I actually tried a workaround using dummy queues that seems to be working, but I have concerns about the impact. Here is what I did:
I route the call to a dummy queue.
Inside that dummy queue's In-Queue Flow, I execute a Transfer to User block directed at the specific agent fetched from my Data Table.
While this achieves the direct transfer, it likely affects reporting accuracy, as the interaction is technically touching multiple queues before reaching the agent.
I am still open to new insights and workaround solutions that might achieve this more cleanly within Standard Routing without compromising the reporting data.
------------------------------
Thanks,
Balaji B
Original Message:
Sent: 02-27-2026 07:27
From: Phaneendra Avatapalli
Subject: Technical Query: Achieving Direct Agent Transfer with ACW using Standard Routing
Hi Balaji,
Under Standard Routing, I don't believe it's possible to truly "pin" a call to a specific user using Architect variables or Transfer to ACD properties. Once the interaction is transferred to ACD, the routing engine immediately evaluates all eligible agents and distributes based on the queue's routing method.
One possible approach while staying within Standard Routing would be to control eligibility instead of trying to override distribution. For example, create a unique skill for that specific agent and require that skill in the initial Transfer to ACD. Since only that agent has the skill, the interaction would route to them first while still preserving the queue's ACW and in-queue experience.
If the agent does not answer, you could then handle voicemail or fallback logic within the In-Queue flow.
Happy to hear if anyone has implemented this differently.