Who we are:

We're Dexibit. We love dinosaurs and data. Our Kiwi AI product helps visitor attractions (museums to zoos to theme parks) make sense of their data and make better decisions. We work with big international names alongside local favourites like MOTAT and Weta. We’ve been widely recognised in industry and picked up a few awards along the way from HiTech, NZIBA, TIN and more. 

Dexibit recently joined accesso, a global technology company serving over 1,000 venues across 30+ countries. That means our platform, our reach, and the complexity of the problems we get to solve just got a lot bigger. 

What we build :

Dexibit brings together our deep domain knowledge, AI-ready data platform, served through an agentic interface. We create value for our customers by addressing non-trivial tasks like strategic modelling, scheduling, operational reporting and analysing visitor feedback. 

Under the hood, the AI layer orchestrates real business logic and APIs: integration packs, custom datasets and benchmarks, context control features, and our query and visualisation infrastructure. 

The agentic interface is where we're investing new feature development. Think streaming, context window management, tool orchestration, sandboxed execution. There's real depth here, and real problems to solve. 

Where you fit :

We're looking for a product-minded software engineer who can shape what we build and how we build it. You'll be in the code every day. But we value engineers who lead naturally: through design decisions, technical direction, mentorship, and bringing the team along. 

You're full-stack, but you lean heavily to one side, whether that's backend data systems or frontend product engineering. With the depth of our problems (and the leverage that frontier LLMs now give us), we believe strong engineers need real depth on one side while being capable and willing to contribute across the stack. 

What you bring :

You break complex problems into simple, elegant solutions. You're a great communicator with strong customer empathy. You're quick to understand the why behind the work. 

You're a proven engineer (5+ years) with experience across different systems and domains. You've shipped impactful work and can talk honestly about the failures as well as the wins. 

You bring curiosity and humility. You mentor naturally. You're not afraid to give constructive feedback, and you genuinely enjoy learning from others' perspectives. 

Our tech stack :

• Languages we use: English, TypeScript, Java, Python, SQL 

• Data systems: PostgreSQL, DuckDB, Databricks, OLTP/OLAP 

• Platforms: cloud native on AWS (with Ansible + CDK), Datadog, Modal, LangFuse 

• AI/LLM: OpenAI/Anthropic models, Pydantic, AI SDK, LangChain4J, XGBoost, Prophet 

• UI: React, Jotai/Redux Toolkit, BFF, data viz, possibly WASM in the future 

• Systems: mix of serverless, services, and monolith 

• Way of working: cycles, communication over process 

• Dev tooling: Claude Code, Codex. We build AI, and we build with AI 

Details :

• Auckland-based, hybrid (3 days in-office / 2 at home) 

• We're open on level. Whether you're a strong senior engineer or someone with 15 years under your belt, we'll meet you where you're at in terms of title and compensation. 

Even if you can't tick every box, we still encourage you to apply. Dexibit recognises all our team as diverse and unique individuals. 

Who we are:

We're Dexibit. We love dinosaurs and data. Our Kiwi AI product helps visitor attractions (museums to zoos to theme parks) make sense of their data and make better decisions. We work with big international names alongside local favourites like MOTAT and Weta. We’ve been widely recognised in industry and picked up a few awards along the way from HiTech, NZIBA, TIN and more. 

Dexibit recently joined accesso, a global technology company serving over 1,000 venues across 30+ countries. That means our platform, our reach, and the complexity of the problems we get to solve just got a lot bigger. 

What we build :

Dexibit brings together our deep domain knowledge, AI-ready data platform, served through an agentic interface. We create value for our customers by addressing non-trivial tasks like strategic modelling, scheduling, operational reporting and analysing visitor feedback. 

Under the hood, the AI layer orchestrates real business logic and APIs: integration packs, custom datasets and benchmarks, context control features, and our query and visualisation infrastructure. 

The agentic interface is where we're investing new feature development. Think streaming, context window management, tool orchestration, sandboxed execution. There's real depth here, and real problems to solve. 

Where you fit :

We're looking for a product-minded software engineer who can shape what we build and how we build it. You'll be in the code every day. But we value engineers who lead naturally: through design decisions, technical direction, mentorship, and bringing the team along. 

You're full-stack, but you lean heavily to one side, whether that's backend data systems or frontend product engineering. With the depth of our problems (and the leverage that frontier LLMs now give us), we believe strong engineers need real depth on one side while being capable and willing to contribute across the stack. 

What you bring :

You break complex problems into simple, elegant solutions. You're a great communicator with strong customer empathy. You're quick to understand the why behind the work. 

You're a proven engineer (5+ years) with experience across different systems and domains. You've shipped impactful work and can talk honestly about the failures as well as the wins. 

You bring curiosity and humility. You mentor naturally. You're not afraid to give constructive feedback, and you genuinely enjoy learning from others' perspectives. 

Our tech stack :

• Languages we use: English, TypeScript, Java, Python, SQL 

• Data systems: PostgreSQL, DuckDB, Databricks, OLTP/OLAP 

• Platforms: cloud native on AWS (with Ansible + CDK), Datadog, Modal, LangFuse 

• AI/LLM: OpenAI/Anthropic models, Pydantic, AI SDK, LangChain4J, XGBoost, Prophet 

• UI: React, Jotai/Redux Toolkit, BFF, data viz, possibly WASM in the future 

• Systems: mix of serverless, services, and monolith 

• Way of working: cycles, communication over process 

• Dev tooling: Claude Code, Codex. We build AI, and we build with AI 

Details :

• Auckland-based, hybrid (3 days in-office / 2 at home) 

• We're open on level. Whether you're a strong senior engineer or someone with 15 years under your belt, we'll meet you where you're at in terms of title and compensation. 

Even if you can't tick every box, we still encourage you to apply. Dexibit recognises all our team as diverse and unique individuals.