Agentic AI Intern

Canda ads; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

Canda ads is seeking engineers to advance production-grade agentic AI systems. You will build integrations across platforms and data sources to form unified agent workflows that communicate in a natural, human-like way across messaging and social channels.

You will work with MCP-based integration frameworks, Hermes/OpenClaw orchestration, and connect services such as Instagram, LinkedIn, Gmail and calendar systems into cohesive pipelines.

Qualifications

• Experience shipping agentic AI systems in production.
• Fluency with MCP standards and connector platforms.
• Comfort across multiple integration frameworks and channels.

Responsibilities

• Integrate platforms and data sources into unified agent workflows.
• Develop agents with natural, human-like communication across messaging channels.
• Build and maintain persistent data layers for agent systems.
• Perform browser automation for apps lacking APIs (e.g., dating apps).
• Connect third-party services into cohesive pipelines (Instagram, LinkedIn, Gmail, calendars).
• Create WhatsApp or Telegram agents with conversational voice and tone.

Skills

Agentic AI systems

MCP standards

Connector platforms

Multi-platform integration

Tools

Composio

Nango

Hermes

OpenClaw

PostgreSQL

Mem0

Job description

Batch: 2025/2026/2027.

About the role:

• Work on production agentic AI systems that have already been deployed and are handling real-world tasks
• Build integrations connecting multiple platforms and data sources into unified agent workflows
• Develop agents that maintain natural, human-like communication across messaging and social channels

What you'll work on:

• Integration work using MCP (Model Context Protocol) and connector frameworks such as Composio, Nango or equivalent platforms
• Regular hands‑on development with Hermes and/or OpenClaw for agent orchestration and execution
• Connecting third‑party services including Instagram, LinkedIn, Gmail and calendar systems into cohesive agent pipelines
• Creating WhatsApp or Telegram agents that engage users with authentic, conversational voice and tone
• Implementing persistent data layers using R2, PostgreSQL, Mem0 or comparable memory systems
• Browser automation for applications that lack formal APIs, including Hinge, Tinder and Instagram DMs

What we’re looking for:

• Shipped experience building and deploying agentic AI systems in production environments
• Strong technical fluency with MCP standards and connector platforms
• Comfort working across multiple integration frameworks and communication channels
• Familiarity with memory and persistence patterns in agent systems