Backend + Prompt Engineer

About The Position

You should be able to show real projects - agent systems you’ve built, code samples, or walkthroughs. You’ll need to explain how your agent handles memory, what tools it uses, and how it recovers from errors.

This is a hands-on, engineering-heavy role. You’ll be building things that go live fast.

What is your new project?

• Domain: AI, eLearning
• Location: Israel
• Company size: 2-10

• Founded in: 2024

   

What makes this project exciting?

Join our client in building a socially responsible AI tool designed to help parents empower their children in an era where digitalization and artificial intelligence shape every aspect of life. This innovative app allows parents to tailor the experience to their child’s unique needs, ensuring a highly personalized approach to learning and self-growth.

What makes this project truly special is its visionary founder—a former Head of Product at Facebook and Instagram. He brings deep industry expertise and a track record of success.

If you’re eager to turn bold ideas into reality, explore the open positions below and be part of something transformative!

What makes you a great fit

You’ve built and shipped agent systems before - not chatbots, but structured agents with state, logic, and integrations. You understand how to make a system that reacts to inputs, plans across turns, and takes meaningful action.

You’re comfortable with:

• Designing prompts that call tools and manage turn-by-turn logic
• Building Python services and deploying to GCP (Cloud Functions, Vertex AI)
• Modeling agent memory (short-term vs persistent state)
• Debugging prompt/tool errors and improving response reliability
• Working with Firestore or similar NoSQL stores for live memory/state

Stack Overview

• Infra: GCP (Vertex AI, Firebase, Firestore, Cloud Functions)
• Backend: Python (FastAPI, async services)
• Agents: Custom-built, using Vertex Agent Builder tool integrations

• Data: Firestore (task state, memory, history), Cloud Storage (media), Google Identity

   

Your day-to-day in this position

You’ll be responsible for designing and building the core AI agents that power the product’s parent-facing system. These agents don’t just respond to prompts - they plan, remember, use tools, and execute structured routines with and for families.

You’ll:

• Design multi-step agent flows using Vertex AI (Tool use, memory, and state handling)
• Write and maintain structured prompts for real-world task execution
• Build and maintain backend services in Python (FastAPI, Cloud Functions)
• Use Firestore and Firebase to store and manage user state and agent memory
• Integrate third-party APIs and internal tools for things like scheduling, content validation, and screen time control
• Work closely with our backend lead to maintain a scalable and observable system

Our next steps

✅ Submit a CV in English — ✅ Intro call with a Recruiter — ✅ Managerial interview — ✅ Client's technical interview — ✅ Offer