AI Agent Developer

We're hiring a full-time AI Agent Developer to build autonomous intelligent systems for our agency and clients. You'll architect, develop, and deploy production-ready agents that make decisions, use tools strategically, and execute multi-step workflows.

What you'll do:
- Design and build autonomous AI agents from scratch
- Architect multi-step workflows with state management
- Implement tool use and function calling with LLMs
- Integrate agents with external APIs, databases, and data sources
- Optimize agent performance, costs, and reliability
- Collaborate directly with founder on technical decisions

This is NOT: Simple API integration work / Building basic chatbots / Tutorial-level projects / Following step-by-step instructions

This IS: Solving complex architectural problems / Making technical decisions autonomously
Building systems that work reliably at scale / Debugging why agents make certain decisions

About MarketioLab: Swedish AI + automation agency serving global companies ($0.5M-$50M revenue). We build practical AI solutions + Agents that drive real business value.

Initial Projects:
1. B2B Prospecting Agent
Autonomous lead generation system that monitors LinkedIn, company databases, and news, researches prospects, qualifies based on criteria, updates CRM with reasoning.

2. Automotive Intelligence Agent
Continuous monitoring system for 50+ car dealership websites, AI-powered vehicle matching against buyer criteria

3. Industry Insights Agent
Multi-source research pipeline that: * Scrapes industry news, blogs, reports, social media
* Identifies trends, competitor movements, market shifts
* Uses RAG to avoid repetition and maintain knowledge base and more

Tech Stack
Core Technologies:
-Python OR TypeScript (strong typing required for production)
-FastAPI (Python) OR Next.js 14 (TypeScript) for APIs
-PostgreSQL with proper schema design for agent state
-Background jobs (Inngest, Celery, BullMQ, or similar)
-Vector databases for RAG (Pinecone, Chroma, Supabase Vector)
-Redis for caching and rate limiting

AI/LLM Integration:
-OpenAI API (GPT-4, function calling, structured outputs)
-Anthropic Claude API (tool use, long context)
-Understanding of: prompt engineering, cost optimization, model selection

Agent Architecture (we'll train on our stack, but you must understand concepts):
-State management patterns (how to persist and resume workflows)
-Tool orchestration (when to use which API/tool)
- Error recovery strategies (retries, fallbacks, graceful degradation)
- Conditional logic and branching (if X fails, do Y)
-Production observability (logging, tracing, debugging)

Preferred Experience (but not required if you have equivalent skills):
- LangChain, LangGraph, Vercel AI SDK, CrewAI, or custom agent frameworks
- Experience with n8n, Make, or similar shows understanding of workflow concepts

Data & Integration: REST APIs and webhooks / Web scraping (Puppeteer, Playwright) as agent tools. / Redis for caching and rate limiting

Requirements -> Must Have:
- 3+ years professional software development
- Built at least one production AI agent (not tutorials/chatbots)
- Hands-on experience with LangChain/LangGraph or similar
- OpenAI function calling or Claude tool use experience
- Strong TypeScript OR Python
- PostgreSQL or similar database experience
- GitHub portfolio with agent code we can review

Work Environment -> Schedule:
Full-time, 40 hours/week / Flexible hours with 2-3h overlap with CET 9am-12pm
Async-first communication culture

Daily Workflow:
Morning Slack standup: Yesterday/Today/Blockers
End of day update: Completed work and progress
GitHub for code reviews and collaboration

Who You'll Work With:
Direct collaboration with founder + developer.
Technical discussions, code reviews, and architectural decisions together.

Tools Provided:
Unlimited OpenAI/Anthropic API credits for development
Any necessary SaaS tools / Development tool subscriptions you need

Compensation & Growth: Starting Salary: $1,400-2,000/month (full-time)

Your specific offer depends on: Demonstrated AI agent experience in portfolio
Quality and approach in paid test task + How quickly you can be productive

Growth Path:
3-month review with raise potential
Top performers: $1,800-2,000+ within 6 months
Opportunity to shape technical direction + Potential profit-sharing as we scale

—-----


How to Apply -> Applications missing these requirements will be auto-rejected

Email: [your email]
Subject: AI Agent Developer - [Your Name]
Must Include:
1. GitHub Portfolio Link (required)
Show us agent/AI code you've written:
LangChain/LangGraph implementations / Function calling or tool use examples / Multi-step workflows with state management / Real code you wrote (not just forked repos)

2. Answer These 3 Questions:
a) "Describe the most complex AI agent you've built"
Include: What problem did it solve? What tools/APIs did it use? How did it d ---------- which tool to use? What challenges did you face? Link to code.


b) "Explain the difference between a chatbot and an agent"
Then explain when you'd use GPT-4 vs Claude Sonnet, and when you'd use streaming vs batch processing.


c) "Link to code using OpenAI function calling or Claude tool use"
Explain what it does and why you structured it that way.

3. Start Your Email With:
"I have [X] years building AI agents with [framework name]"
(This proves you read the full job post :) )


—----

Selection Process
~2 weeks from application to start
Step 1: Application Review (within 48 hours)
Step 2: Technical Screening (30 min video)
Top 3 move to test task
Step 3: Paid Test Task ($150) / Build a "Company Research Agent":
Step 4: Final Interview (30 min)
Step 5: Onboarding
Start within 1-2 weeks