Built in Pakistan. Trusted globally.
We started Poocho AI because enterprise AI products built in California don't understand how people speak in Lahore, Karachi, Riyadh, or Dubai — and because organisations in regulated sectors can't afford to send their data to someone else's cloud.
Why we built this
The contact centres of banks, government agencies, hospitals, and telecom operators across South Asia and MENA handle millions of conversations every day — mostly in Urdu, Arabic, and regional languages that mainstream AI vendors treat as second-class. Agents handle the same routine calls over and over. Wait times are long. Customers hang up frustrated. And the moment a regulator asks where the data went, nobody has a clean answer.
Poocho AI was founded in Lahore in 2024 to solve all three problems at once: native language understanding (not translation), enterprise-grade security and on-premise deployment, and contact-centre automation that actually works in the call types organisations lose sleep over — fraud disputes, citizen complaints, patient appointments, subscriber churn.
We are headquartered in Lahore with a presence in the UAE, and we serve customers across Pakistan, the UAE, Saudi Arabia, and the broader MENA region. Our engineering, NLP research, and customer-success teams are all based in Pakistan — which means when you call our support line, someone picks up in Urdu or English, in your timezone, on the same business day.
The team
Salman leads product vision and enterprise sales at Poocho AI. His background spans machine learning, software architecture, and building technology products for regulated industries across Pakistan and the Gulf.
Fawaz drives the research and architecture behind Poocho AI's language models. He brings deep expertise in NLP, speech AI, and deploying large-scale ML systems in production environments.
Azhar oversees engineering execution and platform reliability. He leads the team building the integration adapters, telephony stack, and real-time speech pipelines that power Poocho AI deployments.
Adnan manages delivery and client operations across Poocho AI's enterprise deployments. With a PMP background, he ensures implementations go live on time and integrations meet enterprise-grade quality standards.
Murtaza leads the core engineering team building Poocho AI's voice and chat infrastructure. He is responsible for platform scalability, security architecture, and the systems that handle tens of thousands of concurrent conversations.
What we believe
Language is not a feature
Urdu and Arabic are not bolt-ons to an English model. They are the primary experience for our customers and their customers. We train for them from the ground up.
Data sovereignty is non-negotiable
Your customer data belongs to you, stays on your infrastructure, and is subject to your regulatory regime — not ours. We have no interest in becoming a data aggregator.
Regulated industries deserve better tools
Banks, hospitals, and government agencies face the most complex AI deployment environments. We build for their constraints first, not as an afterthought.
Automation should feel like service
An AI interaction should be faster, clearer, and more respectful of the caller's time than waiting on hold. If it isn't, we haven't done our job.