Lead / Senior AI Engineer

Nutun; Johannesburg; On-site; ZAR 1,200,000 - 1,800,000; Full time

Job summary

Nutun seeks a hands-on Lead AI Engineer to own delivery of conversational voice and text agents at scale in production. You’ll line-manage a compact engineering team while staying close to the code, acting as a true player-coach reporting to the Chief Analytics Officer.

You will partner with the Lead AI Architect on technical direction, design and build production Python services on Azure/Kubernetes, and lead prompts, evaluation suites, and operational reliability.

Qualifications

• 7+ years software engineering; 3+ years shipping LLM and/or speech+LLM systems to production.
• Deep experience in at least two of: voice/conversational agents, speech pipelines (STT/diarization/VAD), LLM evaluation, multi-agent/tool-calling systems.
• Solid Azure experience (serverless and/or Kubernetes, queues/messaging, blob storage, SQL, secrets/identity).
• Prompt engineering for multi-step, compliance-sensitive workflows.
• Proven player-coach leadership in a small agile team.

Responsibilities

• Own end-to-end delivery for one or more major areas (voice agents, speech/LLM QA pipelines, post-call intelligence, or text-channel agents).
• Partner with the Head of AI on architecture, priorities, and production risk; turn decisions into shipped systems.
• Design and build production Python services on Azure (Functions and/or Kubernetes workers, messaging, storage, SQL-backed outcomes).
• Lead prompt/agent design, structured LLM outputs, evaluation/regression suites, and operational reliability (latency, cost, failure handling, observability).
• Drive post-call intelligence quality: scoring accuracy, structured output reliability, and pipeline performance.
• Mentor Senior Junior engineers through design reviews, pairing, and code review.
• Collaborate with the platform portal team so campaigns, dashboards, and QA surfaces stay aligned with the AI runtime.
• Work with Azure DevOps on deployability, GPU/LLM cost trade-offs, and production operations.
• Help plan and deliver email and WhatsApp conversational agents using shared platform patterns.
• Use and champion AI coding tools (e.g. Cursor, Claude) to raise team speed and quality.

Skills

LLM integration

Voice/pipeline (STT/TTS)

Text agents

Python

API design

GitHub-based workflow

Production AI systems at scale

People leadership

Tools

Azure

Kubernetes

Docker

CI/CD

Hugging Face ecosystem

Eval frameworks

Job description

The role:

Lead a compact AI engineering team hands-on: own the delivery of conversational voice and text agents running at scale in production, line-manage the engineers, and stay in the code. You manage the people while the division’s Lead AI Architect directs the technical approach - a genuine player-coach seat, not a meetings job.

Must have:

Deep LLM integration, voice-pipeline (STT/TTS) and text-agent experience, Python, API design, GitHub-based team workflow, production AI systems at scale, people leadership of small engineering teams.

Nice to have:

Telephony platforms, collections or fintech domain, Azure, Kubernetes/Docker, CI/CD, Hugging Face ecosystem, eval frameworks.

Purpose

Player-coach technical lead who owns delivery of major AI product areas, repoting into the Chief Analytics Officer and/or Lead AI Architect. Ships substantial code, sets engineering standards, mentors the team, and drives production quality across voice AI, automated call QA, and text channels.

Responsibilities

• Own end-to-end delivery for one or more major areas (voice agents, speech/LLM QA pipelines, post-call intelligence, or text-channel agents)
• Partner with the Head of AI on architecture, priorities, and production risk; turn decisions into shipped systems
• Design and build production Python services on Azure (Functions and/or Kubernetes workers, messaging, storage, SQL-backed outcomes)
• Lead prompt/agent design, structured LLM outputs, evaluation/regression suites, and operational reliability (latency, cost, failure handling, observability)
• Drive post-call intelligence quality: scoring accuracy, structured output reliability, and pipeline performance
• Mentor Senior Junior engineers through design reviews, pairing, and code review
• Collaborate with the platform portal team so campaigns, dashboards, and QA surfaces stay aligned with the AI runtime
• Work with Azure DevOps on deployability, GPU/LLM cost trade-offs, and production operations
• Help plan and deliver email and WhatsApp conversational agents using shared platform patterns
• Use and champion AI coding tools (e.g. Cursor, Claude) to raise team speed and quality

Must-Have Skills & Experience

• 7+ years software engineering; 3+ years shipping LLM and/or speech+LLM systems to production
• Deep experience in at least two of: voice/conversational agents, speech pipelines (STT/diarization/VAD), LLM evaluation, multi-agent/tool-calling systems
• Solid Azure experience (serverless and/or Kubernetes, queues/messaging, blob storage, SQL, secrets/identity)
• Prompt engineering for multi-step, compliance-sensitive workflows
• Proven player-coach leadership in a small agile team

Preferable / Strong Advantages

• Experience with AI coding assistants such as Cursor, Claude, or similar
• Voice platform experience and STT/TTS vendor tuning
• Whisper / diarization / vLLM / OpenAI Batch / structured outputs
• Contact-centre, collections, or BPO domain experience
• Observability (Application Insights, Grafana) and CI/CD collaboration