Junior Software Engineer – AI Platform

About the Position

We are building a first-of-its-kind AI-driven integration platform that fundamentally changes how enterprise systems connect. Instead of spending weeks hand-coding data pipelines and digging through messy documentation, our platform uses Large Language Models and agentic AI to generate production-ready integration code directly from meeting notes, emails, and other unstructured documents. Think of it as a CI/CD pipeline for business requirements.

This is a junior role, but deliberately not a narrow one. You will not be handed one layer of the stack and left there. In your first year you will touch React components, backend endpoints, LLM prompts, retrieval pipelines, agent workflows, and tests - often inside the same feature. We build with AI agents as our core engineering tool, not as a side experiment, and that changes what a junior engineer can cover: the agent does much of the typing, so your job is knowing what to ask for, whether the output is actually right, and what it takes to make it production-ready.

What AI will not do for you is decide whether you are building the right thing. You steer. You review. You own whether it is correct when it ships.

What you will help build

• The Brain - an AI knowledge hub that deeply understands complex APIs, data models, and business rules.
• The Sandbox - a web-based IDE where architects and developers collaborate with AI to draft and test data pipelines in minutes instead of weeks.
• The Engine - a fleet of AI agents that keeps integration code up to date in real time as customer requirements change.

Key Responsibilities

Frontend

• Develop and maintain interactive frontend components in React / TypeScript - editors, diff views, run-and-inspect panels, and the surfaces where humans review what the AI produced
• Build UI for AI that behaves like AI: streaming responses, long-running jobs, partial results, and states where the model is uncertain or wrong

Backend

• Create backend API endpoints handling file uploads, AI workflows, and multi-tenant state
• Turn an AI capability into a real feature: the endpoint, the job that runs it, the errors it raises, and the retries when a model call fails

AI and agents

• Integrate LLM APIs (Anthropic, OpenAI) into shipped product features, not demos
• Build LLM-powered utilities that extract structured data from unstructured documents - meeting notes, emails, specifications, PDFs
• Work with vector databases and RAG pipelines: chunking, embedding, retrieval, and honest evaluation of what actually comes back
• Write, test, and iterate prompts and multi-step agent workflows, and debug them when they misbehave

Quality and delivery

• Write and maintain unit tests, plus evaluations for AI outputs - ordinary correctness and “did the model do the right thing” are both your problem
• Debug non-deterministic systems: trace a bad answer back to the prompt, the retrieved context, or the code around it
• Own a task through the full cycle: requirement → implementation → test → review → ship → document, with clean work in Git

Requirements

Technical fundamentals

• Solid programming fundamentals in at least one language - and genuine willingness to work in the others
• Working knowledge of React: components, state, hooks. Enough to build a screen, not only style one
• Ability to build and consume HTTP / REST endpoints, and to work confidently with JSON and schema-shaped data
• Understanding of the full software development lifecycle (requirements → specification → development → testing → deployment → documentation) and clean work in Git
• Experience may come from professional work, internships, personal projects, or hackathons. What matters is that you have built and shipped something real

AI fluency - the part that matters most

• Genuine, hands-on fluency with LLMs and AI agents: writing and iterating prompts, building or configuring agents, orchestrating multi-step workflows
• You steer the agent rather than treating it as a black box. You can explain why an output was wrong and what you changed to fix it
• Healthy skepticism: you verify AI output instead of shipping it because it compiled
• Interest in working with AI agents as your core tool, not a side experiment

Communication and collaboration

• Confident conversational English. You will be in design discussions and demos
• Curiosity to understand a problem before automating it - willing to keep asking “why does it work this way?” until you actually know
• Strong team player: shared backlog, regular syncs, code review in both directions. This is not solo work
• Comfort with ambiguity. This is a young product, and a specification is sometimes a mockup, a meeting transcript, and a conversation
• Hybrid work (office in Gdańsk, Poland) is required

Will be a Plus

• Experience with an agent framework or tooling ecosystem (MCP, LangChain / LangGraph, or similar)
• Experience with a vector database (pgvector, Pinecone, Qdrant, Weaviate, or similar)
• Document parsing and extraction work: PDF, DOCX, email threads, transcripts
• Any prior work with evaluation, tracing, or observability tooling for LLM applications
• Cloud basics: containers, CI/CD, and one of AWS / Azure / GCP
• Bachelor’s degree in computer science or equivalent

Prior experience not required

• No prior AI-platform or enterprise-software background required
• No production LLM experience required, as long as your hands-on AI work is real and you can talk through it

How success is measured

Early on: how quickly you get from “task assigned” to “reviewed, tested, merged” without the work needing to be redone. After that, on two things - breadth, meaning how many parts of the platform you can confidently pick up a task in, and judgment, meaning whether the AI-generated code you approved was actually correct. AI tools support you at every step; the accountability for what ships is yours.

Where this role leads

The role is broad on purpose, because we are growing the engineers who will own the core of this platform. Depending on where you turn out to be strongest, the path from here runs deeper into AI and agent engineering, or wider into full-stack product engineering on the platform itself.

About the Company

ABCloudz is a global full-service company operating at the cutting edge of the latest application development and data management technologies. At ABCloudz, we believe optimal data environments are the greatest competitive advantage, and our Solutions Architects have experience with building enterprise-grade systems.

It is why our team has dedicated ourselves to helping hundreds of organizations – from start-ups to Fortune 100 companies – with architecting cloud solutions, developing web and mobile applications, optimizing complex data workloads, and managing workloads on an ongoing basis - no matter where those datasets may reside.

Growth

Join an always-growing company to operate on the cutting-edge of information technologies, cloud services, application development, and digital infrastructure. At ABCloudz, we created a perfect environment to raise your professional skills, understand new technologies, and work with world-class companies.

Expertise

Our mission is to help businesses accurately, efficiently, and reliably build, move, and manage data workloads. The trusted ABCloudz team now has hundreds of employees in five countries, including our headquarters in Redmond, Washington, and offices in San Francisco, Israel, Poland, and Ukraine.

Mission

Our expertise covers all major cloud and on-premises and databases, including Oracle, SQL Server, Azure, AWS, and GCP solutions. Our prolific developers are focused on delivering the best possible digital solutions leveraging our expertise with BI, Advanced Analytics, AI, Big Data, IoT, DevOps, and Managed Services.

Your dream job not found?

We always keep the doors open for outstanding talents. Want to work with us, but couldn't find an appropriate opening? Let us know about your passion, tell us why you stand out from the crowd and how you can make the difference working at ABCloudz. Just hit the button below to send us your resume.

Send Resume