Loading
Open to select work · Q3 2026/Islamabad ⇔ Everywhere
AI Engineer · Data Scientist
HaiderFarooq.
I engineer the systems that replace teams. Core engineer on TryCook.ai's growth OS and Aligno's dual-portal clinic platform — production AI that compresses weeks into minutes and scales without headcount.
Islamabad, PK
Portrait '26
$12.1M+Generated
348Founders
5,500+Patients
141Districts
01Selected Work
02Capabilities
I build the kind of systems that used to require entire teams.
From multi-agent AI operating systems that compress 5–8 weeks of work into 90 minutes, to dual-portal clinic platforms serving 5,500+ patients, to civic infrastructure that covers all 141 districts of Pakistan — the common thread is scale without headcount, and craft without shortcuts.
I work across the full stack: agentic backends, production frontends, data pipelines, and deployment. One engineer, end-to-end ownership.
01
Agentic AI systems
Multi-agent orchestration, LLM cost engineering, evals, and chained-agent product design.
02
Production frontends
Marketing surfaces, dashboards, dual-portals — designed and shipped end-to-end.
03
Data + MLOps
Spark/Kafka pipelines, drift-aware retraining, fairness audits, and streaming analytics.
04
Civic infrastructure
National-scale tools that move bureaucracy — 141 districts, 1,000+ officials, zero friction.
03The work, in numbers
$12.1M+Client Revenue
348Founders on Platform
5,500+Patients Served
141Districts Covered
4AI Products in 1 Year
90 minDown from 5–8 Weeks
04About
“Engineering should feel like writing — clear, intentional, and yours.”
Haider Farooq · Islamabad
I’m Haider — an AI engineer and data scientist based in Islamabad, currently finishing a B.S. in Data Science at NUCES while shipping production systems for clients across North America, Pakistan, and Australia.
My work lives at the intersection of agentic AI and product engineering. I don’t just connect LLM APIs — I design the workflows, the cost architecture, the frontend, and the deployment. From a $12.1M growth OS to a civic hazard map covering an entire nation, the scale is always outsized for the team size.
Next.jsTypeScriptPythonPyTorchPySparkApache KafkaApache SparkApache StormHadoop / MapReducePostgreSQLMongoDBDistilBERTHugging Facescikit-learnXGBoostSHAPFairlearnAIF360KubeflowMLflowFastAPIDockerKubernetesPrometheusGrafanaTailwindFramer MotionGHLOpenAI / Claude / GeminiAgentic AILLM orchestrationCost-aware billing
05Get in touch