Role Specification

Senior Full - Stack Engineer - Applied AI

Location: South Africa / Remote arrangement

Employment type: Full - time

Experience: 6+ years

Build AI systems that survive contact with reality

We are building a cloud-based, multi-tenant application that turns frontier AI models into practical tools for real organizations.

We need a senior full-stack engineer who can architect, build, deploy, and operate complete product capabilities. You will work across the user experience, application services, data, AI orchestration, cloud infrastructure, security, and production operations.

This is not an academic, model-training, or traditional MLOps role. We are not looking for someone whose primary experience lives in notebooks, PyTorch, or research experiments. We want a product-minded software engineer who knows how to make sophisticated technology useful - and get it into customers’ hands. Our stack is predominantly TypeScript and Azure, but it evolves continuously. You must be comfortable learning rapidly, challenging current assumptions, and adopting better approaches as the technology changes.

What you will build:

At the centre of our product is a distributed AI runtime: a full-stack harness that coordinates frontier models, application state, context, tools, data, and cloud services.

You will help design and build:

• Responsive, streaming AI experiences using React, Next.js, and TypeScript.
• Cloud services that manage conversations, context, model calls, tools, subagents, and long-running work.
• Tool and MCP integrations that let models work safely with enterprise applications and data.
• Reliable orchestration across OpenAI, Anthropic Claude, and other frontier-model providers.
• Retrieval and data experiences spanning structured data, documents, search, embeddings, and business systems.
• Authentication, authorization, tenant isolation, spending controls, and auditable tool access.
• Resilient execution, including timeouts, retries, fallbacks, cancellation, failure recovery, and model migrations.
• Production infrastructure, deployment automation, monitoring, and operational tooling on Azure.

You will not be confined to one layer. On the same capability, you might shape the interaction design, implement a React component, build a streaming API, improve tool execution, change a persistence model, configure an Azure service, and diagnose its production behaviour.

What you will own:

• Take ambiguous customer and product problems from discovery through production.
• Make architectural decisions across frontend, backend, data, AI, and cloud infrastructure.
• Build maintainable product capabilities rather than stopping at proofs of concept.
• Explore new model capabilities and turn the useful ones into dependable features.
• Understand how OpenAI and Claude models behave when using tools, following complex instructions, handling context, and recovering from failure.
• Integrate enterprise data sources without compromising security or tenant isolation.
• Improve latency, reliability, scalability, observability, and cost.
• Ship through automated, testable, and reversible deployment processes.
• Diagnose difficult problems that cross application, model, identity, networking, and infrastructure boundaries.
• Help establish engineering standards and raise the technical capability of the team.
• Measure success through customer and business value-not novelty or technical theatre.

What we are looking for:

• Six or more years of professional experience across full-stack software engineering, cloud application development, analytics engineering, data platforms, or production systems.
• A strong software engineering background: you have built and operated scalable, customer-facing applications-not just prototypes or notebooks.
• Deep TypeScript and Node.js skills, with strong experience in React and modern web application development.
• The ability to work across the entire system: user experience, APIs, application services, data, integrations, cloud infrastructure, deployment, and production operations.
• Substantial Azure experience, or sufficiently strong cloud fundamentals to become effective in Azure quickly.
• Practical experience with identity, permissions, secrets, networking, observability, security, and multi-tenant application design.
• A strong understanding of modern applied-AI systems: frontier models, agents, tool use, MCP, context engineering, retrieval, streaming, reliability, latency, and cost.
• Recent hands-on experience building applications with OpenAI, Anthropic Claude, or comparable frontier models is highly beneficial.
• An understanding of evaluation, tracing, regression testing, and model lifecycle management. Direct production experience in these areas is valuable, but this is not an MLOps specialist role.
• A record of taking ambiguous technical problems from exploration to production and measurable user value.
• The judgement to distinguish a genuine capability advance from a fashionable demo.
• Clear written and verbal communication. You can make complex technical trade-offs understandable without hiding behind jargon.

A degree is welcome but not required. Evidence wins.

You do not need experience with every technology in our stack. We care more about engineering depth, intellectual curiosity, learning speed, ownership, and what you have actually shipped than an exact keyword match.

Our technology:

Our technology changes as better capabilities emerge. Today, the environment includes:

• TypeScript, Node.js, React, Next.js, and streaming APIs.
• OpenAI, Azure OpenAI, Anthropic Claude, Amazon Bedrock, and other frontier-model providers.
• Agent orchestration, tool calling, MCP, retrieval, subagents, and enterprise connectors.
• Azure App Service, Cosmos DB, Azure SQL, Azure AI Search, Storage, Redis, Key Vault, Application Insights, and Log Analytics.
• Microsoft Entra ID, Microsoft Graph, SharePoint, and enterprise delegated authorization.
• Databricks, Snowflake, and other organizational data sources.
• Terraform/OpenTofu, Docker, GitHub Actions, deployment slots, private networking, managed identities, and WAF.

Python experience is useful, but this is primarily a full-stack product engineering role.

We expect you to understand systems, not merely recognize technology names. We also expect the list to change.

How you work:

You are curious, direct, demanding of yourself, and comfortable being accountable for outcomes.

You study new models, development tools, and engineering techniques because you need to understand what has become possible-not because someone assigned you a ticket. You prototype quickly, measure honestly, discard weak ideas without sentiment, and harden the ideas that work.

You are equally comfortable working on a user interaction, a TypeScript service, a data model, an AI tool, or an Azure deployment. You do not retreat into one layer of the stack or declare difficult cross-cutting problems to be someone else’s responsibility.

You care about architecture, security, reliability, and maintainability, but you do not use them as reasons to avoid shipping. You know that the work is only complete when it is creating value in production.

You are obsessed with the edge of what technology can do, but practical about applying it. Novelty earns attention; demonstrated value earns investment.

This is a high-ownership, high-intensity environment. If you want fixed scope, a comfortable pace, and work that stays neatly inside a job description, this will not be the right role.

If you want unusual autonomy, difficult problems, direct responsibility, and the opportunity to push frontier AI into serious production use, we should talk.

We value sustained performance and good judgement-not theatre, unnecessary hours, or reckless shipping.

Show us what you have built

A strong application will include specific examples of:

• A full-stack cloud application or major capability you personally helped take to production.
• The parts you owned across architecture, implementation, deployment, and operation.
• A technically difficult production failure you diagnosed.
• A time you used an emerging technology to deliver measurable customer or business value.
• How you currently use OpenAI, Claude, or AI-assisted engineering tools in your own work.
• Something you tried that did not work-and what the evidence taught you.

Please tell us what you personally owned, what failed, what you measured, and what outcome changed.

About us:

Amplify is an enterprise AI company helping organisations apply AI securely across real business workflows. Our core platform enables businesses to create, manage and scale purpose-built AI agents that connect with their existing knowledge, data and systems. It provides the security, governance, visibility and cost control required for confident enterprise adoption, while supporting leading AI models through a single, flexible platform.

We serve medium-sized organisations and large-enterprises across industries, particularly those with complex operations, valuable knowledge assets and stringent security requirements. Amplify addresses the gap between experimenting with generative AI and delivering it safely at scale - helping customers turn fragmented information and manual processes into governed, repeatable AI-enabled workflows that improve productivity, decision support and operational performance.

To apply:

Email recruitment@amplify360.ai with your CV and anything you believe we should see (e.g. repos, projects, cover letter or otherwise)