flowchart LR
G((Goal)) --> P[Planner / Router]
P --> K[Context / RAG]
P --> A[Specialist Agents]
K --> A
A --> T[Tools / APIs]
T --> O[Observation]
O --> M[(Memory)]
M --> V[Evaluator / Critic]
V -->|refine| P
V -->|validated| X[Autonomous Action]
X --> D[Production System]
D --> E[Telemetry / Evaluation]
E -. feedback .-> P
classDef input fill:#0F172A,stroke:#22D3EE,color:#E2E8F0,stroke-width:2px;
classDef brain fill:#1E1B4B,stroke:#818CF8,color:#F8FAFC,stroke-width:2px;
classDef agent fill:#2E1065,stroke:#A78BFA,color:#F8FAFC,stroke-width:2px;
classDef memory fill:#164E63,stroke:#67E8F9,color:#F8FAFC,stroke-width:2px;
classDef action fill:#052E16,stroke:#34D399,color:#ECFDF5,stroke-width:2px;
classDef eval fill:#3F1D2E,stroke:#F472B6,color:#FDF2F8,stroke-width:2px;
class G input;
class P,K brain;
class A,T agent;
class O,M memory;
class X,D action;
class V,E eval;
|
PhD — AI & Autonomous Systems SEECS · NUST AI research at the intersection of agentic systems, autonomous decision-making, multimodal intelligence and robotics. Multi-Agent Systems · VLMs · AutonomyPlanning · Reasoning · EvaluationRobotics · Human-in-the-Loop
|
6+ years · 100+ production projects Engineering intelligent products from architecture and APIs to mobile/web, cloud infrastructure, deployment and observability. Python · TypeScript · FastAPI · Node.jsReact · Next.js · React NativeMongoDB · PostgreSQL · AWS · Docker
|
Representative production work · Explore full DevZona portfolio →
🧠 Evala.aiPitch-deck analysis, scoring and AI-assisted decision workflows. |
Generative AI Document intelligence and founder-facing AI analysis workflows. |
AI Privacy SaaS Automation-oriented privacy and data-removal workflows. |
📱 SafegramReal-time product, privacy systems and backend architecture. |
Logistics Platform Customer, operational and backend systems for fuel delivery. |
🤝 MentrMentor–mentee matching and organizational workflows. |





