class PinkalMistry:
role = "Engineering Manager → Generative AI & Agentic Systems"
location = "Dubai, UAE 🇦🇪"
experience = "12+ years Engineering | 4+ years People & Platform Leadership"
current = "CAFU App DMCC — On-Demand Fuel Delivery Platform"
ai_focus = [
"LLM-Powered Product Features",
"Agentic Workflows & Tool-Calling Pipelines",
"RAG Systems & Knowledge Retrieval",
"Model Context Protocol (MCP)",
"Multi-Agent Orchestration",
"AI-Powered Operational Analytics",
]
stack = {
"LLMs" : ["Claude", "GPT-4o", "Gemini", "Llama 3", "Mistral"],
"Frameworks": ["LangChain", "LangGraph", "CrewAI", "AutoGen"],
"RAG" : ["Pinecone", "ChromaDB", "FAISS", "Weaviate"],
"Mobile" : ["Kotlin", "Swift", "Jetpack Compose", "Android", "iOS"],
"DevOps" : ["GitHub Actions", "Bitrise", "Fastlane", "Firebase"],
}
metrics = {
"crash_rate_reduction" : "8%",
"release_breakage_reduction": "38%",
"customer_reactivation" : "30%+",
"decision_speed_improvement": "40%",
"annual_value_unlocked" : "Multi-million AED",
}
currently_exploring = ["MCP Servers", "Agentic DevEx Tooling", "Platform Observability"]
open_to = ["AI Engineering Manager", "Senior Engineering Manager", "Head of Engineering"]
def mission(self):
return "Ship AI-powered products that move metrics. Lead teams that build with conviction."I'm an Engineering Manager at the intersection of platform engineering and Generative AI. Over 12 years, I’ve grown from building products hands-on to leading cross-platform teams — and today I'm deeply invested in bringing LLMs, agentic workflows, and AI-powered tooling into real-world production systems.
- 🧠 Agentic AI — Systems where LLMs plan, call tools, and execute multi-step tasks autonomously
- 🔗 RAG Pipelines — Retrieval-augmented generation for grounded, production-grade AI features
- 📱 Mobile DNA — Android & iOS roots. I still read PRs and care deeply about app performance
- 🏗️ Platform Thinking — BOT team models, CI/CD at scale, DevEx that ships faster
- 🎯 Impact Obsessed — I measure everything. 8%, 38%, 30%, 40% — numbers are how I tell stories
🤖 Large Language Models
🔗 Agentic Frameworks & Orchestration
📚 RAG & Vector Stores
🛰 Observability & Evaluation
📱 Mobile Engineering
⚙️ Platform, DevOps & Tooling
| 🤖 AI-Powered Analytics | 📉 Crash Rate | 🚀 Release Quality | 💰 Customer Recovery |
|---|---|---|---|
| 40% faster decisions | 8% reduction | 38% fewer breakages | 30%+ reactivation |
| Multi-million AED saved | Across Android & iOS | Via automated CI gates | Via AI re-engagement |
🧩 Agentic Pipelines
├── Tool-calling agents with planning + execution loops
├── Multi-agent systems (CrewAI, LangGraph, AutoGen)
└── MCP server integrations for contextual AI workflows
🔍 RAG Systems
├── Document ingestion + chunking strategies
├── Hybrid search (dense + sparse retrieval)
└── Re-ranking and answer grounding
📊 AI-Powered Operational Intelligence
├── Real-time dashboards backed by LLM summarisation
├── Anomaly detection with natural-language alerts
└── Decision-support systems for operations teams
📱 On-Device & Mobile AI
├── AI-assisted UX flows and smart suggestions
├── Crash analytics augmented with AI root-cause analysis
└── LLM inference integrated into mobile product features
✅ AI-powered Operational Analytics → 40% faster decisions, multi-million AED saved
✅ BOT team scale-up 0 → 15 engineers → full knowledge transfer, 38% fewer breakages
✅ AI-driven customer re-engagement → 30%+ reactivation rate
🔄 In Progress
├── Agentic workflow platform for internal ops automation
├── Multi-agent orchestration with LangGraph + MCP
└── On-device LLM inference for mobile features
🎯 Next Up
├── Open-source agentic AI toolkit for mobile engineering teams
├── RAG-powered developer knowledge base
└── Writing & speaking on AI in Engineering Leadership
open_to:
roles:
- Software Engineering Manager / Senior Software Engineering Manager
- Head of Engineering
domains:
- AI-native Product Companies
- Mobile-first Platforms
- Developer Tooling & Infrastructure
location: Dubai 🇦🇪 | Open to Remote 
