MULTI-AGENT SYSTEMSAGENTIC AIRAG PIPELINESLLM ENGINEERINGAUTONOMOUS WORKFLOWSTOOL-USE AGENTSVECTOR DATABASESLANGCHAINOPENAIFASTAPIREASONING SYSTEMSMCP PROTOCOLFINE-TUNINGPYTORCHHUGGING FACEMULTI-AGENT SYSTEMSAGENTIC AIRAG PIPELINESLLM ENGINEERINGAUTONOMOUS WORKFLOWSTOOL-USE AGENTSVECTOR DATABASESLANGCHAINOPENAIFASTAPIREASONING SYSTEMSMCP PROTOCOLFINE-TUNINGPYTORCHHUGGING FACEMULTI-AGENT SYSTEMSAGENTIC AIRAG PIPELINESLLM ENGINEERINGAUTONOMOUS WORKFLOWSTOOL-USE AGENTSVECTOR DATABASESLANGCHAINOPENAIFASTAPIREASONING SYSTEMSMCP PROTOCOLFINE-TUNINGPYTORCHHUGGING FACE
Available for AI Roles

I engineer
Agentic AI
systems that reason, act, and scale.

AI Engineer specializing in agentic systems, autonomous workflows, and multi-agent architectures, with strong foundations across LLM applications, generative AI, and end-to-end AI engineering.

View Work Contact
Muhammad Dayyan
Muhammad DayyanAI ENGINEER
AI AGENTS
VOICE AGENTS
AUTOMATION
MULTI-AGENT
AI SOLUTIONS
// about me

AI Engineer focused on architecting high-impact AI systems across agents, voice, and automation. I deliver robust, scalable solutions built for real operations and long-term growth.

📥Input
🧠Planner
🔍Retriever
💾Memory
⚙️Executor
🛠️Tool Use
🔄Self-Correct
✅Output
📥Input
🧠Planner
🔍Retriever
💾Memory
⚙️Executor
🛠️Tool Use
🔄Self-Correct
✅Output
📥Input
🧠Planner
🔍Retriever
💾Memory
⚙️Executor
🛠️Tool Use
🔄Self-Correct
✅Output
// services

What I offer.

I provide AI engineering services that help businesses launch practical, production-ready systems — from intelligent agents and voice agents to workflow automation and end-to-end AI delivery.

AI Agent Systems

Designing intelligent AI agent systems that reason, plan, use tools, and execute tasks autonomously across real business workflows.

  • Single-agent and multi-agent architecture design
  • Tool-use integration with APIs, databases, and internal systems
  • Memory, context handling, and decision loops with guardrails

Voice Agent Development

Building real-time voice agents for support and operations with natural dialogue, context awareness, and action execution.

  • Real-time speech pipeline (ASR + LLM + TTS)
  • Multilingual voice flows and context-aware call logic
  • Low-latency deployment with streaming integrations

AI Workflow Automation

Turning repetitive business processes into AI-powered autonomous workflows that reduce manual effort and improve speed.

  • Workflow orchestration with event-triggered task routing
  • Human-in-the-loop checkpoints where needed
  • Monitoring, retry logic, and continuous optimization

End-to-End AI Solutions

Delivering complete AI solutions from strategy and architecture to deployment and iteration for product and enterprise use cases.

  • LLM apps, RAG systems, AI copilots, and applied AI features
  • Scalable backend services, APIs, and deployment pipelines
  • Performance, cost, and quality optimization for outcomes
// career trajectory

Where I've delivered.

Lead AI Engineer
GrowthRune
May 2025 → Present
Dubai · Remote
  • Leading architecture and delivery of end-to-end agentic AI platforms with autonomous task-executing agents and multi-agent orchestration
  • Designing reasoning-driven workflows with planning, tool use, memory, and self-correcting execution loops
  • Implementing enterprise-grade retrieval systems with hybrid search, optimized indexing, and contextual memory pipelines
  • Building MCP-driven agent communication for dynamic routing, shared context, and interoperable tool execution
  • Shipping scalable AI microservices and event-driven automation engines integrated with internal and external APIs
AI Engineer
Octaloop Technologies
Dec 2024 → Apr 2025
Islamabad
  • Built AI applications across conversational interfaces, computer vision pipelines, and multilingual real-time voice systems
  • Implemented agentic task automation and retrieval-augmented AI for context-aware enterprise use cases
  • Applied efficient fine-tuning approaches to adapt LLMs for domain-specific performance and lower compute cost
Generative AI Intern
DevSort
Jul 2024 → Nov 2024
Islamabad
  • Developed retrieval-enhanced chatbot systems with external knowledge grounding for more accurate responses
  • Worked on diffusion-based generation workflows and efficient model adaptation for high-quality output
// tech stack

What I work with.

🤖
Agentic AI
Autonomous Systems
🧩
Multi-Agent
Coordination & Routing
🧠
Reasoning Workflows
Planning + Tool Use
💬
LLM Engineering
Prompting · Evaluation
🔍
RAG Systems
Hybrid Retrieval
✨
Generative AI
Text · Image Workflows
🎙️
Voice Agents
Real-Time AI Assistants
🔊
Speech AI
ASR · TTS · Turn Handling
📞
Voice Infrastructure
WebRTC · WebSockets · Streaming
🗣️
NLP
Conversational Intelligence
👁️
Computer Vision
Classification · Detection
🐍
Python
Core Language
⚡
FastAPI
AI Services
🔗
LangChain
Orchestration
🔧
OpenAI SDK
Model Integration
🤗
Hugging Face
Models & Pipelines
🔥
PyTorch
Deep Learning
📊
TensorFlow
ML Framework
🐳
Docker
Containers
🔄
CI/CD
Deployment Automation
🗄️
Vector DBs
Semantic Search
// selected work

What I've built.

All projects under NDA — capability-focused summaries.
Agentic AI Platform

Autonomous Multi-Agent Orchestration Engine

Built a production agentic platform where specialized agents decompose goals, select tools, share memory, and execute multi-step workflows through iterative reasoning and self-correction loops.

Agentic AIMulti-AgentLangChainOpenAIPython
Confidential
Knowledge Intelligence

Enterprise Retrieval & Context Engine

Developed an enterprise-grade retrieval architecture combining hybrid search, contextual memory, and response-grounding pipelines to improve answer quality, relevance, and trust in high-volume Q&A workflows.

RAGVector DBEmbeddingsRetrievalPython
Confidential
Voice Agent AI

Real-Time Multilingual Voice Agent Platform

Engineered low-latency voice agents for live customer interactions, integrating speech recognition, contextual dialogue management, and multilingual response generation for natural, action-oriented conversations.

Voice AIASR/TTSLLMFastAPIWebSockets
Confidential
Generative AI

Efficient Diffusion Fine-Tuning Pipeline

Built a diffusion training pipeline with lightweight adaptation methods to reduce compute overhead while preserving output fidelity, enabling faster iteration and domain-specific image generation.

DiffusionLoRAHugging FacePyTorch
Confidential
// let's connect

Let’s build
autonomous AI that delivers outcomes.

Open to full-time roles, contract engagements, and collaborations in Agentic AI, multi-agent systems, LLM applications, and end-to-end AI product engineering.

m.dayyaann@gmail.com
Email GitHub LinkedIn Call