Arda Cezzar
Chief executiveSets company direction and carries the commercial side of every customer and partner agreement.
Team
Built in Istanbul, live in nine countries
Four crafts share one room: product, AI labs, backend and client operations, with a named owner on every line.
DRING was founded in 2024 and now runs production calls in nine countries from one Istanbul office.
Who steers
Six people, five clear remits
Nobody here owns everything. Each remit has a named owner end to end, and international business has two: one opens markets, one works with partners and customers abroad.
Ayberk Bozkuş
Product and operationsDecides what gets built and how it runs once it is live, and signs off every release.
Can Mergen
TechnologyOwns the engineering stack, from SIP and carrier routing up to the agent runtime and the API.
Dr. Osman Kahraman
AI labsLeads prompt and evaluation engineering, and the research into models DRING can run on premise.
Hande Kocagöz
International businessOpens markets outside Turkey and carries a first conversation through to an agreed workflow.
Hakan Yazgan
International businessWorks with partners and customers abroad, and brings their workflows back to the product team.
How the work divides
Four crafts, one handover chain
Each craft finishes a piece of work and hands the next one something it can act on.
Product
Turns a customer workflow into a brief a builder can act on: what the agent must settle, what it must never say, and what counts as done. Hands the brief to AI labs.
AI labs
Writes the prompt, the guardrails and the evaluation set, then runs the simulated conversations that decide whether the agent is ready. Hands a tested agent to backend.
Backend
Owns the path the voice actually travels: SIP trunks, carriers, media, the CRM write-back and the API. Hands a connected line to client operations.
Client operations
Runs the line after launch: watches live calls, reads the scores, reports weekly and writes down what the next release has to fix. Hands the evidence back to product.
The release loop
A brief goes round, and comes back scored
The same loop runs for a first agent and for the fifth one on an account that is already live.
The brief
Product writes down the job, the boundaries and the outcome the agent has to leave behind.
Simulated first
AI labs runs 1,000 to 10,000 simulated conversations, built for that company, before any real call.
On a real line
Backend puts the agent on carrier routes and wires the write-back into the customer's own systems.
Watched live
Client operations reads the calls and the scores while the line runs, not in a report a month later.
Back to the brief
What the calls showed becomes the next release, and the loop starts again with a tighter brief.