Introduction Our client is looking for the following candidate to join a vibrant Platform Development team of 25 people in an intellectually stimulating environment to solve complex challenges on their Software Developer platform, SCP.apps.
As a senior full-stack developer with an agentic focus, you will help re-architect the SCP.apps platform to be AI-native by design, shipping AI agents and LLM-powered features that form the heart of the product, internal tooling, and the shared agent platform.
Who This Role Would Suit:
This role is tailor-made for an exceptionally skilled Full Stack JavaScript Developer who has transitioned into the cutting edge of AI engineering. It is perfect for a senior developer who goes far beyond casual prompt use and has real-world experience shipping LLM-powered features, building autonomous agent workflows, and managing production realities like latency, non-determinism, and token budgets.
You are a self-directed professional who thrives in an intellectually stimulating environment, capable of influencing project direction without direct command when translating unclear requests into highly structured, scalable code. You have a natural curiosity for emerging AI tools and models, combined with the pragmatic engineering discipline required to build safe, deterministic guardrails and long-running agent memory networks. If you want to move away from basic application wrappers and help re-architect a massive enterprise developer platform to be truly AI-native by design, this is your next challenge.
Full Stack Mastery: 8+ years of deep expertise in full-stack JavaScript/TypeScript development (ES6+, Node.js, React, GraphQL) delivering high-performing, scalable platforms and features. AI & Agentic Engineering: Proven production track record of embedding LLMs, designing multi-step agent workflows (tool/function calling, structured outputs), managing context windows, and building retrieval-augmented generation (RAG) pipelines. Technical Leadership: Act as a completely independent functional expert to break down complex requirements, shape target AI architectures, mentor junior engineers, and champion secure, responsible AI boundaries.
Position Details:
Contract Dates: 01-09-2026 to 31-12-2029 Location: Midrand / Menlyn / Rosslyn / Home Office Rotation Role Group: DevOps Role Name: Software Engineering Level: 6-8 Years related experience
Duties & Responsibilities Key Responsibilities:
AI-Native Architecture: Shape the target architecture for SCP.apps, deciding where agents and retrieval should replace or augment existing flows, and help lead the migration. Product AI & Tooling: Re-architect core product flows around AI agents, ship customer-facing features (in-product copilots, automations), and build internal agents to optimize testing and operations. Agent Platform Development: Develop and maintain shared agent building blocks—including tool integrations, MCP servers, and data retrieval (RAG)—for other engineers to build upon. Full-Stack Engineering: Develop full-stack features using Node.js, Apollo, React, and GraphQL, implementing reusable UI components, authentication, and permission models. Quality, Safety & Production: Establish evaluation suites, tracing, and guardrails (cost control, data privacy, prompt-injection defenses). Own model routing, token optimization, and non-deterministic behavior handling. Agile & Roadmap: Project manage via epics/user stories, participate in code reviews, run proof-of-concepts on emerging models, and help drive the platform roadmap. Mentorship & Support: Guide and mentor junior developers and provide community technical support (2nd and 3rd level on a rotational basis).
Important Application Details
Location & Relocation
Applicants based outside of Gauteng must be willing to relocate. Please note that relocation to the province will be at the candidate's own cost.
Eligibility & Legal
Citizenship: South African citizens and residents are preferred. Work Permits: Candidates with valid work permits will be considered. Privacy: By applying, you consent to being added to our database and receiving updates until you unsubscribe.
Application Status
If you do not receive a response within 2 weeks, please consider your application unsuccessful.
#isanqa #isanqaresourcing #fuelledbypassionintegrityexcellence #thegroup #FullStackDeveloper #SeniorDeveloper #TypeScript #NodeJS #ReactJS #AIEngi
Desired Experience & Qualification Qualifications & Experience:
Relevant IT Degree and/or equivalent professional qualifications. 8+ years of commercial experience in full-stack JavaScript development or a similar role. Demonstrable experience designing, building, and shipping LLM-powered platforms or agentic features to production. Strong understanding of SDLC methodologies (Agile). Essential: A clear habit of staying current in a fast-moving field, evaluating new models, tools, and techniques pragmatically. Advantageous: AI or ML coursework or certifications.
Essential Skills & Technologies:
Core Tech Stack:
TypeScript / JavaScript (ES6+) Node.js (Express.js, Apollo) React (with TypeScript) & Webpack GraphQL & REST APIs Databases: MongoDB, PostgreSQL & Redis Docker & Docker Compose, Git
AI & LLM Production Expectations:
Hands-on experience embedding LLMs into production applications. Building agentic features: tool and function calling, structured outputs, and multi-step agent workflows. Applied prompt engineering and context management (system prompts, token/context-window management). Working knowledge of retrieval-augmented generation (RAG). Sound judgment on LLM realities: managing latency, cost, token budgets, and non-determinism.
Advantageous Skills:
AWS cloud experience (specifically ECS, Lambda, and AWS Bedrock or SageMaker for model hosting). Multi-agent orchestration and agent-to-agent patterns (e.g., planner/executor, supervisor/worker). Building and exposing Model Context Protocol (MCP) servers and tools. Agent evaluation, tracing (quality, cost, safety), and observability metrics. Guardrails engineering (permission scoping, human-in-the-loop approvals, audit trails). Agent memory/state management for long-running tasks via Redis or Vector stores. Model selection, smart routing across providers, and prompt/response caching. Vector databases and embedding-model selection.
Interested? Explore Our Latest Opportunities: iSanqa Careers Page
Stay in the Loop: Subscribe to our weekly mailing list so you don’t miss out!
Join Our Job Notification Channels: Stay updated on WhatsApp or choose your specific field on Telegram:
WhatsApp Channel | IT Roles (Telegram) | SAP Roles (Telegram) Pharmaceutical & Medical (#) | Finance (#) | Manufacturing & Engineering (#) Supply Chain & Procurement (#) | Sales & Marketing (#)
Communication Preferences: Please let us know how you would like us to manage your details:
Opt-In: I want to receive updates on new opportunities via mail, call, or text. Database Only: Keep my profile on file, but do not contact me. Remove: Completely remove my details from the database.
Sourced from external listing
iSanqa
Sourced from PNet