Founding Software Engineer - Agentic Systems

At Veriff, we are exploring the next frontier of identity verification: agentic AI systems that can reason, act, and operate safely in high-trust environments.

We are forming a small, high-leverage team to explore and productionize agentic capabilities across identity, KYA, risk assessment, and automation workflows.

We are looking for a Founding Engineer to help build this from the ground up.

You’ll help us protect honest people online by:

• Design and build agentic systems that can:
  • Use tools;
  • Reason across multi-step workflows;
  • Operate under strict compliance and safety constraints.
• Prototype and productionize LLM-powered workflows;
• Build internal evaluation frameworks for agent reliability;
• Define system architecture for scalable, secure AI services;
• Work closely with product, risk, and compliance teams;
• Turn ambiguous ideas into working systems.

You are the right future Veriffian for the job if you have:

Strong Systems Engineer

• 6+ years building backend systems in production
• Strong Python experience (FastAPI or similar)
• Experience with distributed systems and async workflows
• Deep understanding of system design and reliability

Agentic / LLM Experience

Hands on experience building:

• Tool using agents
• Multi step reasoning workflows
• Retrieval Augmented systems (RAG)
• Multi agent or orchestration systems

Experience with one or more:

• LangChain / LangGraph
• OpenAI Assistants or Responses API
• AutoGen
• CrewAI
• Semantic Kernel

You have shipped LLM-powered systems beyond experimentation.

Elements of Our Tech Stack

• Backend: Node.js (TypeScript), Python, Golang
• Frontend: React (TypeScript)
• Mobile: Kotlin (Android), Swift (iOS), Dart (Flutter), TypeScript (React Native)
• Infrastructure: AWS, Kubernetes, Terraform
• Data & ML: PostgreSQL, RabbitMQ, Python-based ML models

Nice to have:

• Experience in fraud, risk, compliance, or identity systems
• Experience building internal AI evaluation pipelines
• Familiarity with vector databases (pgvector, Pinecone, etc)
• Experience operating in regulated environments

What Success Looks Like

• Reliable agentic workflows running in production
• Clear evaluation metrics for AI quality and safety
• Strong architectural foundations for scaling AI capabilities
• High quality technical documentation and system clarity