Skip to main content
Sami Ullah

Sami Ullah

Agent Solution Architect · Multi-Agent Systems · Scalable AI Infrastructure

Talent
HaripurMember since July 16, 2026
LangGraphYOLOMLflowDVCFastAPIPostgreSQLRedisGitHub ActionsNext.jsReactSQLBackend DevelopmentLangchainAgentic WorkflowsMiddleware ArchitectureHuman-in-the-LoopMulti-Query ExpansionQuery DecompositionRerankingpgvectorSemantic SearchAutonomous Web Search AgentsPythonDockerAlembicDocker ComposeCI/CD
·

Hire this person

Send a job offer directly to this candidate

LangGraphYOLOMLflowDVCFastAPI

About

AI Engineer with 2+ years designing, shipping, and reviewing production-grade agentic systems. Founder of ConnectXeo — a multi-agent Shopify platform where a coordinator delegates to 7 subagents across 40+ live tool calls. Author of langctl (published on PyPI, 14+ releases) — an open-source CLI that scaffolds, deploys, and standardizes LangChain agent projects with built-in middleware, memory, and one-command deployment.

Track record of defining agent orchestration patterns, architectural standards, and scalable multi-tenant infrastructure.

Experience

Founder & Agent Architect - ConnectXeo

(2025-01)

  • Architected the multi-agent system from scratch — coordinator-subagent pattern with 7 specialized agents, each owning isolated tool scopes, error boundaries, and retry logic across 40+ live Shopify API integrations
  • Defined and enforced architectural standards for agent-to-agent communication, state management, and tool-call routing — reusable patterns now governing every subsystem on the platform
  • Reviewed agent design decisions across multiple project domains (orders, refunds, inventory, draft orders) — ensuring each subagent scales independently behind consistent contract interfaces
  • Built multi-tenant infrastructure with per-merchant agent isolation, shared orchestration patterns, and multi-channel deployment (WhatsApp Business Cloud API, web chat, REST API)

AI/ML Engineer - Revnix (rext.ai)

(2025-03)

  • Sole backend architect on a production content automation platform — designed multi-stage agent orchestration: automated topic discovery → relevance scoring → LLM outline generation → human-in-the-loop approval → SEO-optimized publishing
  • Engineered async state persistence with PostgreSQL + Alembic migrations and Redis-based workflow orchestration, ensuring pipeline reliability under production load

AI Engineer - Emulxion.ai

(2024-01 - 2024-11)

  • Designed advanced RAG architecture with multi-query expansion, query decomposition, fusion, and intelligent reranking — established the retrieval pattern adopted across the engineering team
  • Built real-time inference pipeline serving production traffic via React + FastAPI with sub-second response times

Education

BS - Computer Science - University of Haripur (2022 - 2026)

Skills

Agent Solution ArchitectMulti-Agent SystemsScalable AI Infrastructure

Reviews

Related