AI/ML Engineer (Agentic AI & Vision-Language/LLM Integration)

Company Overview

Sigmawave AI is a visual synthetic data company helping organisations build better computer vision and AI systems. We provide configurable, high-quality synthetic datasets that support model training, testing, and validation—delivered safely, efficiently, and at scale.

Job Description

We are hiring an

AI/ML Engineer (Agentic AI + Vision/LLM Integration)

to integrate existing

Vision LLMs

and

Large LLMs

into

Terra Builder , enabling AI-assisted generation of production-ready 3D assets from visual inputs and intelligent, prompt-driven creation and editing workflows. You will focus on model integration, orchestration, evaluation, and deployment—working closely with product and engineering teams to deliver reliable capabilities to end users.

Responsibilities

Agentic AI & Orchestration

Design and implement

agentic AI workflows

that translate user intent into system actions using

tool/function calling , structured outputs, and robust execution logic. Build reliable

planning and execution loops

(multi-step plans, retries, safe fallbacks, validation, and rollback strategies). Develop mechanisms for

precise targeting and scope-limited edits

(e.g., editing only selected regions or attributes) while maintaining predictable behaviour.

Vision/LLM Integration for Asset Generation

Integrate existing

Vision LLMs and LLMs

(open-source and/or commercial) into the platform to support

AI-assisted 3D asset generation and editing

from user-provided visual references. Implement end-to-end integration patterns: request/response schemas, queuing, validation, conversion/packaging, and user-facing outputs. Collaborate with 3D/graphics and platform teams to ensure outputs meet platform requirements (format compatibility, asset validity, and performance constraints).

Quality, Evaluation & Optimisation

Define evaluation methods for quality and consistency (e.g., visual fidelity, intent adherence, repeatability, and failure mode tracking), including optional human-in-the-loop review flows. Optimise inference for

latency, cost, and reliability

(caching, batching, quantisation where relevant). Build monitoring and diagnostics for model behaviour in production (telemetry, logging, alerts, quality regressions).

Productionisation & Platform Readiness

Package AI capabilities as

APIs/microservices , ensuring secure integration, observability, and maintainability. Maintain clear documentation: integration guides, configuration, and operational runbooks.

Required Qualifications

3+ years delivering ML systems into production (or equivalent depth via shipped projects). Strong Python skills; practical experience with

PyTorch

and modern model tooling (e.g., Hugging Face ecosystem). Experience integrating and operating

LLMs

and

multimodal / vision-language

models via APIs or self-hosting. Agentic AI experience : tool/function calling, structured prompting, planning/execution loops, reliability guardrails, and evaluation approaches for agents. Familiarity with production deployment patterns (REST/gRPC), Docker, and MLOps fundamentals (CI/CD, monitoring, versioning). Strong problem-solving skills and ability to work cross-functionally in a product environment.

Nice to have Experience with 3D pipelines or 3D ML (asset formats such as OBJ/FBX/GLB, UVs, materials, mesh processing). GPU inference optimisation experience (ONNX Runtime, TensorRT, vLLM, quantisation frameworks). Familiarity with retrieval systems (vector DBs), agent evaluation harnesses, or orchestration frameworks (LangGraph/LangChain/LlamaIndex).

What We Offer

High ownership over a core AI capability in a production platform. Work on cutting-edge

multimodal + agentic AI

applied to real customer needs. Collaborative team environment across product, platform, and 3D/graphics engineering. Competitive compensation (commensurate with experience). Flexible working arrangements (subject to team needs).

How to Apply

Email

with: Your CV/resume Links to GitHub/portfolio and relevant demos (agents, multimodal systems, model serving, 3D asset pipelines) A short note describing 1–2 similar systems you’ve built and what you owned end-to-end Email subject:

Application – AI/ML Engineer (Agentic AI & Vision-Language/LLM Integration)

Reference:

SWA-AIML-VISION-AGENTIC-001