Ergün Mercan
speaker
129 appearances
1 recordings
1 series
first heard Jul 2026
last heard 13 Jul
Ergün Mercan’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.
Appearances
So the real win is for the remaining human agents who are now completely freed from repetitive low-value tasks, which could be things like password resets or basic modem setups.
So they finally have the time to focus deeply on complex customer queries that require genuine human problem solving.
By utilizing the domain-specific agents that are hard-coded with telecom standard operating procedures and vocabularies, we have built a highly predictable automated asset for our customer that functions as a permanent infrastructure layer for continuous cost reduction.
No, absolutely.
So humans are central to this strategy.
The industrial data we track shows that while 60 to 70 of highly repetitive data-driven telco cases can be handled entirely by autonomous agents, the remaining 30 to 40 absolutely require human expertise.
This is what we call human in the loop.
These are, for example, complaints representing serious churn risk, emotional or high value interactions, or even complex technical cases.
These will never be fully automated and the human touch will always be necessary.
So the goal isn't to replace humans, but to augment them.
This is why we design for highly structured semi-autonomous deployments or human-agent collaboration workspaces.
It's a persistent operational environment that shifts the human workforce from manual delivery to high-level supervision and stewardship.
The workspace embeds explicit confidence thresholds.
When an autonomous agent encounters a highly complex technical edge case, a severe billing dispute, or a high-value interaction representing a serious churn risk, it triggers a seamless contextual handoff.
So the human manager acts as the ultimate authority, reviewing the agent's decision traces, adjusting the intent, and managing the exceptions safely.
It's a true partnership that plays directly to the unique strengths of both machine skill and human empathy.
So first, you must establish a clear financial framework for agentic AI from day one.
Traditional software has always been highly predictable, linear, per user licensing costs or static costs, but autonomous agent architectures do not.
It's because multi-agent orchestration involves recursive reasoning loops.
It could be tool calls, automated retries.
Showing 81–100 of 129 · page 5 of 7
← Previous
Next →