Software Engineer
First seen by Alion on Sep 8, 2026.
About the role :
As a Software Engineer on the Intelligence team, you will work across the systems that turn raw sensor readings into decisions : data pipelines processing telemetry from thousands of pieces of equipment, machine learning models that predict how that equipment behaves, and an AI agent that answers operators' questions from live building data.
This is a deliberately broad role. In your first year you will touch data engineering, applied machine learning and backend work, and you will find out which of them you enjoy most. You will work alongside senior engineers who review your code and explain their reasoning, and with product managers and customers who will tell you directly whether what you built is useful.
We do not expect you to know anything about buildings or HVAC when you start. Most of us did not.
Tech Stack :
- Languages : Python (primary), JavaScript, TypeScript
- Machine learning : scikit-learn, XGBoost, LightGBM, Keras, pandas, numpy, AWS SageMaker
- Reinforcement learning and control : Model Free RL, Deep RL, ML based Optimiser
- AI systems : AWS Bedrock, tool-calling agents, retrieval-augmented generation
- Data platform : Apache Airflow, InfluxDB, PostgreSQL, Redis, S3
- Knowledge graph : Amazon Neptune, Brick schema
- Observability : Grafana, Prometheus, Loki, Sentry, OpenTelemetry
- Backend and front end : FastAPI, Node.js, Angular
- Edge : ARM Linux, MQTT
- Cloud and infrastructure : AWS, Docker, GitHub Actions
What you will do :
- Explore and clean real sensor data - the kind with gaps, drift, stuck values and units that disagree - and work out what it is actually telling you.
- Build and maintain data pipelines that turn raw equipment telemetry into the metrics our models and our product depend on.
- Train, evaluate and retrain machine learning models on plant sensor data - power draw, water flow, temperatures, and the efficiency baselines used to verify savings.
- Write tools for our AI agent : functions that fetch data from our knowledge graph and time-series stores and return it in a form a language model can reason over.
- Trace data problems across the stack, from a sensor on a machine through message transport, storage and APIs to the dashboard someone is looking at.
- Write tests, particularly for the cases that have already caught us out once.
- Demonstrate your work - to the team regularly, and to customers as you grow into it.
- Visit sites. Seeing the equipment your code affects will change how you build for it.
You may be a good fit if you have :
- Comfort with Python, and the ability to read and modify code you did not write. Familiarity with pandas and numpy.
- Statistics fundamentals - distributions, mean against median, variance, correlation, outliers, and the difference between a correlation and a cause.
- Exploratory data analysis as a habit.
- Machine learning fundamentals from projects you have built end to end.
- Familiarity with scikit-learn. Exposure to XGBoost, LightGBM, or a deep learning framework such as PyTorch, TensorFlow or Keras is a bonus.
- Working knowledge of SQL and basic data modelling.
- Comfort with Git, the command line and a Linux environment.
- A habit of checking your own work.
- Willingness to explain what you built to people who do not write code.
- Curiosity about a domain you do not yet know.
Nice to have :
- Any project built with large language models - retrieval-augmented generation, an agent with tool use, a chatbot over your own documents.
- Neural networks and deep learning, in any framework, applied to any problem.
- An understanding of what reinforcement learning is and where it fits.
- Open source contributions of any size.
- Projects involving time-series data or IoT.
- Exposure to AI coding assistants such as Claude Code, Codex, Cursor or Copilot.
- Docker, and any cloud platform.
- Any hardware or control-systems project, academic or personal.
- Experience working directly with users or customers, in any role.
Logistics :
- Minimum education : Bachelor's degree, or an equivalent combination of education, training and experience.
- Experience : 0 - 2 years. Recent graduates with strong internships or substantial personal projects are encouraged to apply.
Skills
Python, Javascript, TypeScript, AWS, Machine Learning, Deep Learning, Knowledge Graph, Data Pipeline