class WasifSohail:
name = "Wasif Sohail"
location = "Rawalpindi, Pakistan"
roles = ["Tech Lead", "DevOps Architect", "Agentic AI Pioneer"]
companies = [
"Co-Founder @ ZEERO AI β zeeroai.tech",
"Tech Lead @ CallRolin",
"AI/ML Engineer @ Devrolin"
]
education = "BS Artificial Intelligence | CGPA: 3.95 / 4.0"
expertise = {
"agentic_ai" : "LangChain Β· LangGraph Β· CrewAI Β· AutoGen Β· Multi-Agent Systems",
"llm_rag" : "GPT-4o Β· Gemini 1.5 Β· LLaMA Β· RAG Pipelines Β· Vector DBs",
"voice_ai" : "Voice Agent @ Devrolin Β· Urdu & Arabic TTS Β· Speech Recognition",
"cv_nlp" : "YOLOv8 Β· MediaPipe Β· OpenCV Β· Transformers Β· BERT Β· HuggingFace",
"fullstack" : "Next.js Β· FastAPI Β· Django Β· TypeScript Β· Python Β· React",
"devops" : "AWS Β· GCP Β· Docker Β· Kubernetes Β· GitHub Actions Β· CI/CD",
"data_ai" : "DataFlow AI Β· Pandas Β· Auto-Visualization Β· NL-to-SQL",
}
currently_building = [
"Voice Agent System @ Devrolin β production-grade AI calling",
"ZEERO AI β Multi-agent system for 3D real-time lectures",
"CallRolin V2 β Urdu & Arabic TTS, STT, and Local LLMs",
"PrismX Extension Host β AWS DevOps infrastructure",
"Advanced RAG with hybrid search & context compression",
]
contact = "wasifsohail66@gmail.com"
fun_fact = "I automate things so I can focus on harder problems"| Domain | Stack |
|---|---|
| π€ Agentic AI | LangChain Β· LangGraph Β· CrewAI Β· AutoGen Β· n8n |
| π LLMs & RAG | GPT-4o Β· Gemini 1.5 Β· LLaMA Β· Pinecone Β· ChromaDB |
| π£οΈ Voice & Speech AI | Voice Agents Β· CallRolin Urdu/Arabic TTS Β· Speech Recognition Β· Azure OpenAI |
| ποΈ Computer Vision | YOLOv8 Β· OpenCV Β· MediaPipe Β· CNNs Β· ONNX |
| π£οΈ NLP | Transformers Β· BERT Β· HuggingFace Β· NER Β· Classification |
| π₯ Healthcare AI | Disease Prediction Β· Patient Clustering Β· Medical Imaging |
| π Data AI | NL-to-insight platforms Β· Auto-visualization Β· Predictive Analytics |
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β WASIF'S DEVOPS STACK β
β βββββββββββββββ¦ββββββββββββββββ¦βββββββββββββββββ¦ββββββββββββββββββββ β
β βοΈ CLOUD β π¦ CONTAINERSβ π CI / CD β π OBSERVABILITY β
β AWS Β· GCP β Docker β GitHub Actionsβ Logging Β· Metrics β
β Azure Β· DO β Kubernetes β Jenkins β Alerting Β· Tracingβ
β Lambda Β· EKSβ Compose β ArgoCD β Performance Monit.β
ββββββββββββββββ©ββββββββββββββββ©βββββββββββββββββ©ββββββββββββββββββββ β
Languages
AI Β· ML Β· Deep Learning
Fullstack & Deployment
Cloud & DevOps
Databases Β· Vector Stores
Production-grade AI voice calling agent at Devrolin. Handles real-time conversations, intent detection, and agentic decision-making β deployed at scale for enterprise clients.
Engineered localized Urdu and Arabic Text-to-Speech (TTS V2) engines, highly accurate Speech-to-Text models, and a localized LLM built completely from scratch locally.
A complete multi-agent ecosystem designed specifically for 3D real-time interactive lectures. Orchestrates multiple AI agents to revolutionize immersive learning environments.
Built and open-sourced memoryOS, a highly optimized software extension engineered to be deployed natively at the edge for advanced contextual memory management.
LLM-powered marketing automation β AI content generation (blogs, ads, emails), campaign management, audience segmentation, competitor analysis. Live on Vercel.
Stack: Next.js + FastAPI + GPT-4 + RAG
Enterprise data platform β upload any dataset, get AI-generated visualizations, ask questions in natural language, auto-correlation analysis, outlier detection.
Stack: Next.js + FastAPI + Python
AWS-deployed extension hosting infrastructure for PrismX β scalable DevOps setup with cloud deployment, multi-cloud infrastructure management, and production-grade hosting pipeline.
Implementation of "Comprehensive Study on Weapon Detection". Extends computer vision research with ONNX export for ultra-fast edge device deployment in smart surveillance systems.

