Lead AI Engineer (Agentic Systems)
About Viseera
Viseera builds AI technology for the Australian vocational education and training sector. Our work gives training organisations the ability to operate at a scale that manual processes cannot reach.
We are an early stage company. We are currently funded by our Founder and underpinned by four established operating businesses within the EFLR Group, and we are moving toward our first seed round. Those sister companies, including one of Australia’s largest private construction training organisations, are our reference customers, our testing ground and our proof that what we build works in production.
That is an unusual position for a startup. We have real users, real revenue and real regulatory constraints from day one, without the pressure of finding a market first.
Why this role exists
We want to build enterprise grade agents that take over low level administrative work so people can do the work that actually matters. Not chatbots. Agents that do the job.
You will work directly and daily with the Founder, and alongside a small technical team that you will guide. Part of the role is lifting the capability of the people around you, setting the standard for how we design and build, and making sure the team is pulling in one direction.
The other part is strategy. You will help shape what we build, what we buy, where AI genuinely creates advantage, and where it does not. If you want a role where the technical decisions and the commercial decisions are made in the same room, this is it.
The role
You will design, build, deploy and maintain production agents end to end, taking ownership of architecture, compute, integrations, security, monitoring, updates and long term maintainability. Human oversight remains part of the design wherever compliance requires it.
You will also be the person who can put a design on a whiteboard, explain it to non-technical colleagues, and then go and build it with the team.
We have a budget to invest in technology and hardware. If the right tool, platform or machine will make the work better or faster, we will back it.
A day in the life
Whiteboarding concepts and problems with the Founder and the team
Designing an agent on paper before a line of code is written
Building agents that solve real operational problems inside the group
Guiding and reviewing the work of the wider technical team
Testing, evaluating and hardening what you built last week
Pitching an idea, catching someone else’s, and moving on it the same day
We run a genuine startup culture. Fast, collaborative and high trust. We pitch and catch. We build, break, learn and rebuild. That is why this is an office role. The work happens in the room, with the team, in person. There is flexibility when you need it, but this is not a work from home position.
Technology environment
Models and platforms
Claude (Anthropic), ChatGPT, Gemini, Grok and Perplexity
Lovable and comparable rapid build platforms
Agentic frameworks and orchestration platforms. We are platform agnostic.
Integrations
HubSpot, ActiveCampaign, Xero, Trello, Shopify, Microsoft 365 and a range of learning management systems
Microsoft Outlook, including reading, interpreting and responding to email
REST APIs, OAuth and webhooks across the group’s systems
Infrastructure
Supabase (database), Vercel (web hosting), Fly.io (compute), cron jobs (scheduled tasks), GitHub (source control)
Key responsibilities
Design, build, deploy and maintain production agentic systems end to end
Guide and develop a small technical team, and set the standard for how work is designed and built
Own architecture and compute decisions, including local versus hosted, resource sizing, inference cost, latency and scaling
Build and maintain integrations across the group’s systems, and keep them running
Establish evaluation and testing practice so agent quality is measured, not assumed
Implement monitoring, logging and alerting appropriate to LLM based systems
Manage model, dependency and platform updates without breaking production
Design for security and privacy from the outset, including handling of student personal information
Govern cost, including token budgets, per agent unit economics and cost visibility
Document designs and explain them clearly to non-technical colleagues before building
Advise the Founder on AI strategy, tooling, build versus buy, and technical risk
Selection criteria
Essential
Demonstrated experience delivering agentic systems in production: OpenClaw, Claude Agent SDK and MCP, LangGraph, CrewAI, Microsoft Agent Framework, n8n or comparable platforms. We are platform agnostic. Sustained production experience matters more than any particular toolset.
Strong API and integration capability: REST, OAuth, webhooks, rate limiting, retry logic and error handling.
Sound judgement on compute and hardware: local versus hosted deployment, resource sizing, inference cost, latency and scaling.
A security first approach: credential and key management, permissions, sandboxing, audit logging, and handling of personal information consistent with the Privacy Act 1988 and the Australian Privacy Principles.
Awareness of adversarial risk in agentic systems: prompt injection, untrusted input handling and tool use containment, particularly where agents read external email.
Operational discipline: version control, evaluation and testing, monitoring, model and dependency updates, and documentation.
Leadership capability: the ability to guide a small technical team, review their work, and raise the standard of what the team produces.
Clear communication: the ability to document a design and explain it to non-technical colleagues before building it.
Seniority and judgement: the confidence to disagree with the Founder, and the evidence to back it.
Desirable
Experience in a regulated industry, or with compliance constrained workflows
Exposure to the vocational education and training sector or another Australian regulated services environment
Early stage or founding team experience
Full stack capability across TypeScript, Python and modern web deployment
What we offer
Competitive market salary, superannuation and entitlements, or a market contract rate if that suits you better
Direct, daily access to the Founder, and genuine influence over technical strategy
A new A grade office in Toowong Tower
Parking available on site, Toowong Village shopping centre directly below, and the train station and bus interchange at the building
Real production workloads and real users from your first week
A small, capable team, a culture built on trust, pace and ideas over hierarchy, and work that is genuinely fun to do
Engagement
We are open on structure. This role can be filled as a full time permanent position, or as a fixed term contract for an agreed period, including a contract that converts to permanent by mutual agreement. Tell us which suits you and we will work with it.
Either way, the expectation is the same. Full time hours, on site in Toowong, working with the team.
How to apply
Apply with a short covering note and links to agentic systems you have shipped, whether that is repositories, demos, architecture write ups or a description of what you built and what broke. We are more interested in what you have put into production than in a formal résumé.
Applicants must have full working rights in Australia.