AI Engineer

vFairs; Pakistan; On-site; PKR 4,000,000 - 6,000,000; Full time

Job summary

vFairs is seeking an experienced AI Engineer to design and ship autonomous AI agents within our virtual events platform. You will work at the intersection of product code and AI agent infrastructure, building tools that automate event setup and orchestration.

You will join a small, high-leverage AI team responsible for agent capabilities and safety, with impact on customer workflows and platform UX. Strong Python, LangGraph, and FastAPI skills are essential.

Qualifications

• 6+ years of professional software engineering experience.
• Hands-on production experience building with LangGraph and FastAPI.
• Experience designing multi-step or multi-agent systems, including task planning and recovery.
• Strong Python skills with LLm-powered applications, agents, RAG, tool-calling, and structured outputs.
• Experience with MCP or similar tool/agent-to-application integration patterns.
• Experience with vector databases (pgvector, Chroma) and SQL-based retrieval strategies.
• Solid understanding of LLM application architecture and production readiness.
• Judgment about autonomy and when to seek human confirmation.

Responsibilities

• Design and build autonomous AI agents that plan, sequence, and execute multi-step workflows for events.
• Architect multi-agent systems with planner/executor patterns and sub-agent delegation.
• Extend and maintain MCP server as the main interface for event data and agent capabilities.
• Design guardrails, validation, and human-in-the-loop checkpoints for safe autonomous actions.
• Build observability pipelines to track agent behavior, drift, and latency for iteration.
• Own the full lifecycle of Agentic features from prototype to production monitoring.
• Collaborate with product to translate customer intents into reliable agent workflows.
• Work with core engineering (PHP/Laravel/Vue) to integrate agent services into the main app.

Skills

LangGraph

Python

FastAPI

Multi-agent systems

LLM-powered apps

Tool orchestration

Autonomy design

MCP protocol

Tools

pgvector

Chroma

SQL

FastAPI

Job description

Job Description:

About the Role

vFairs is looking for an experienced AI Engineer to help us build and scale AI-powered features within our virtual events platform. Youll work at the intersection of our core product and our AI agent infrastructure — designing, building, and shipping intelligent features that make events more engaging and easier to run for organizers and attendees alike.

Youll join a small, high-leverage AI team at the center of that shift — the team responsible for the agents and AI-driven tooling behind our MCP server and support experiences today, and for the autonomous, customer-facing agents were building next. This isnt a team bolted onto the side of the product: the agents you build will become one of the primary ways customers interact with vFairs, doing in minutes what currently takes an event organizer hours of manual setup.

What Youll Do

• Design and build fully autonomous AI agents that can independently plan, sequence, and execute the multi-step workflows required to build out a customers event — from initial setup through configuration of sessions, booths, speakers, and content — with minimal human intervention

• Architect multi-agent systems using LangGraph, including planner/executor patterns, task decomposition, sub-agent delegation, and recovery logic for handling partial failures or ambiguous instructions mid-task

• Extend and maintain our MCP (Model Context Protocol) server as the primary interface through which these agents read and write vFairs event data — defining new tools, resources, and permissions as agent capabilities grow

• Design the guardrails, validation, and human-in-the-loop checkpoints needed to let agents act autonomously on customer data while keeping actions safe, reversible, and auditable

• Build robust evaluation and observability pipelines for agent behavior — tracking task success rate, drift, hallucinated actions, and cost/latency — and use them to drive iteration

• Own the full lifecycle of Agentic features: from prototyping new capabilities, to hardening them for production, to monitoring how they perform against real customer events

• Collaborate with product to translate "build my event" as a customer intent into a reliable, decomposed agent workflow — balancing autonomy against predictability and user trust

• Work with our core engineering team (primarily PHP/Laravel/Vue) to integrate agentic services into the main application, including how agent actions surface and are reviewed in the UI

• Stay current with the fast-moving agent ecosystem (frameworks, evaluation techniques, autonomy patterns) and bring in whats genuinely useful — skepticism of hype is as valuable as enthusiasm for new tools

Requirements

What Were Looking For

• 6+ years of professional software engineering experience

• Hands-on production experience building with LangGraph (or comparable agent orchestration frameworks) and FastAPI

• Experience designing multi-step or multi-agent systems — not just single-turn chat/RAG — including task planning, tool orchestration, and error recovery

• Strong Python skills, with experience designing and shipping LLM-powered applications (agents, RAG, tool-calling, structured outputs)

• Experience with MCP (Model Context Protocol) or similar tool/agent-to-application integration patterns

• Experience with vector databases (e.g., pgvector, Chroma) and SQL-based retrieval strategies

• Solid understanding of LLM application architecture: prompt design, retrieval strategies, evaluation, and handling model limitations/failure modes in production

• Judgment about autonomy: knowing when an agent should act independently versus pause for human confirmation, and how to design systems accordingly

• Comfort working across the stack — from API design to deployment — and shipping features end-to-end

• Strong communication skills and ability to work cross-functionally with product and non-AI engineering teams

Nice to Have

• Experience with PHP, Laravel, and Vue.js — our core application stack

• Experience with multi-cloud infrastructure

• Background in building customer-facing chatbots or support automation

• Experience with WebSocket streaming for real-time AI interactions

Requirements: