About Hachiko

Join us in building the software platform that's accelerating the renewable energy transition, one battery at a time. We're unlocking the C&I sector's storage potential through a combination of energy expertise, mathematical optimisation, software engineering excellence, and an unwavering commitment to customer success.

Our vision is a world where human thriving and a flourishing natural environment are not mutually exclusive. To get there, we empower battery developers, owners, and traders to unlock the full potential of energy storage portfolios and deliver market-leading returns.

We're a small, high-calibre technical team building distributed, event-driven systems that simulate, operate, and optimise real-world energy systems. We value epistemic humility, fast feedback, and continuous improvement. We build production ML systems for energy forecasting and battery optimisation that directly impact customer returns. And we're looking for someone to join our data science team and help us deliver with excellence.

The time to act is now. Come and join us.

About the role

You'll be our second data scientist, working directly alongside our Head of Data Science to accelerate delivery of AI and optimisation features across Hachiko's platform.

This is a high-autonomy, high-impact individual contributor role where you'll work on both production features and exploratory AI capabilities. Your mission is to build models and optimisation models that directly improve battery returns while helping us unlock the next generation of intelligent features.

This isn't a pure research role, nor is it pure production engineering. You'll work on critical forecasting and optimisation problems that drive revenue today, while carving out time to investigate how emerging AI techniques can make our platform smarter tomorrow. We need someone who can ship rigorously and explore creatively.

Compensation will be in the form of a generous base salary, superannuation, and a very generous early-employee ESOP allocation.

What you'll work on

Forecasting & Optimisation: Build and refine time-series models (prices, solar, load) that power trading; enhance battery scheduling under real-world constraints; engineer features from market, weather, and grid signals; run hypothesis tests to validate gains; own the full ML lifecycle from experimentation to monitored production.

Exploratory AI Research: Investigate LLMs, RL, MCPs, and advanced forecasting to unlock new capabilities; prototype intelligent insights on battery performance, market opportunities, and system health; apply AI to optimisation, anomaly detection, and decision systems; balance exploration with shipping.

System Integration: Use SQL to shape time-series data; instrument; create dashboards for monitoring model performance and data quality; collaborate to embed models into an event-driven architecture; think end-to-end so forecast accuracy translates into optimisation outcomes and business impact.

You'll have significant autonomy to shape what you work on, balanced with the reality that we have customers depending on reliable forecasts and optimisations. We're looking for someone who can prioritise ruthlessly and deliver incrementally.

What you'll work with

ML & data science: Python, OpenSTEF, XGBoost, PyTorch, DuckDB, Marimo notebooks, and more

Optimisation: Mixed-integer programming with SCIP

Data & infrastructure: Tiger Data, SQL, Kafka, Grafana, and more

Engineering: Docker, GitHub, CI/CD via GitHub Actions, AWS

We don't expect you to know all of this. Strong fundamentals in time-series modelling, statistics, Python, and SQL are essential. Optimisation experience is highly valuable but we can teach the specifics.

About you

We're after people who embody our values: brevity, empathy, honesty, decisiveness, excellence, and humility.

In summary, we're looking for someone who:

• Fits our culture (effective, curious, humble. Not a kn*bhead, not a hero)
• Aligns with our mission (renewable energy transition matters to you)
• Has the technical chops (proven cloud/platform engineering experience)

You're probably not a great fit if:

• You need perfect data and extensive resources before you start building.
• You optimise for academic rigour over business impact, or vice versa—you can't balance both.
• You see "research" and "production" as incompatible mindsets.
• You need a large team, close supervision, or well-defined playbooks to be effective.
• You're uncomfortable with the messiness of real-world ML: missing data, shifting requirements, integration challenges.
• You're not genuinely interested in renewable energy or climate impact.
• You're looking for a cruisy gig.

Must-Haves

• 3+ years of production ML or applied data science experience: You've built models that run in production and create measurable value, not just research prototypes.
• Time-series forecasting expertise: Deep experience with forecasting techniques such as gradient boosting (XGBoost, LightGBM), deep learning approaches, or classical methods. You understand seasonality, autocorrelation, and temporal validation strategies.
• Statistical foundations: Solid grounding in hypothesis testing, statistical inference, experimental design, and data analysis. You know when to use which test and how to interpret results correctly.
• ML workflows: You understand the full pipeline: data quality assessment, feature engineering, model selection, validation strategies, performance monitoring, and retraining.
• Python proficiency: You write clean, maintainable Python. You understand software engineering basics: version control, testing, code review, documentation.
• SQL competence: You can write complex queries to extract and transform time-series data efficiently.
• Systems thinking: You understand how models fit into larger systems. You reason about error propagation, latency constraints, and production trade-offs.

Strong Preferences

• Optimisation background: Experience with linear programming (LP) or mixed-integer programming (MIP). This is hard to find but exceptionally valuable for our work.
• MLOps experience: Model versioning, experiment tracking, automated retraining, monitoring, A/B testing in production.
• Energy industry exposure: Background in energy markets, grid operations, renewables, or related domains.
• Grafana or similar: Experience building dashboards for monitoring model performance and data quality.
• Early-stage experience: Comfort working in scale-up environments where you define the playbook rather than follow it.
• Research mindset: Genuine curiosity about emerging AI techniques and how they might apply to energy problems.

What we offer

Impact

• Build ML systems that directly accelerate grid decarbonisation and battery adoption
• Work on both production features and exploratory AI capabilities that shape Hachiko's future
• Shape our data science roadmap and practices as an early hire with significant influence

Team & culture

• High-trust, high-agency environment. We hire adults and treat them like adults.
• Work with a small, world-class team that values thoughtful design and collaboration.
• Culture grounded in epistemic humility, fast feedback, and continuous improvement.
• Direct collaboration with domain experts in energy markets, optimisation, and software engineering.
• A team with domain expertise across all areas of the product, true traction with customers, and a huge addressable market.

Growth & learning

• Broad surface area: forecasting, optimisation, feature engineering, MLOps, exploratory AI research, MCPs and data infrastructure.
• Learn about energy markets, grid operations, and battery economics from people who live and breathe it.
• Room to experiment with emerging AI techniques and explore their application to energy problems.
• Balance production delivery with research exploration—develop both skill sets.
• As our second data scientist, you'll have outsized influence on our technical direction.

Work environment

• Primarily remote with regular in-person sessions in Sydney. We care about outcomes, not face time. For the right candidate, we will accept applications from outside Sydney, but expect in-person contact at a reasonable agreed frequency.
• Flexible arrangements: This is a full-time role. However, we are open to part-time options for the right candidate. We value impact over rigid schedules.
• Climate tech impact: Work on technology that directly accelerates grid decarbonisation and the renewable energy transition.
• Small team, high impact: Low bureaucracy, fast decisions, and direct influence on product and technical direction.

How to apply

To apply, send the following to [email protected] with the subject “Data Scientist - <Your name>”:

👉 Your CV

A short paragraph answering the following question: What’s something unique about you that’d make you a better colleague than anyone else?

A Loom video (max. duration 5 minutes) answering the following question:

➡️ “A data scientist on Hachiko's team added new 'grid congestion' features from AEMO (the Australian energy market operator) to our price forecasting model. Results on the test set improved dramatically and the RMSE dropped 40%. But when you deploy it to production, performance is only slightly better than the old model. You have 48 hours to figure out what went wrong. What are your top 3 hypotheses and how would you test them?”

If you’ve never used Loom before, get yourself a free account and record away! Remember to keep the duration to <5 minutes, and make sure you update the link sharing settings (instructions here 👉🏾 https://lnkd.in/g25AWahx).

We value clarity and brevity. If you can't explain your work simply, we're probably not a fit.

Note - applicants must have full rights to live and work in Australia.