This episode’s guest is Tim Rawling, a Founder at CalcTree, a calculation management platform now about four years old. He was on the podcast a couple of years ago, not long after he started there, when the product was positioned as calculation tool that had interesting collaboration and management capabilities. Here we are, two years later and he’s covered a lot of ground since then.

Tim is based in Australia and spends a lot of time talking to global engineering firms about the same set of problems. He is also still on the tools, pushing code and testing workflows, which makes this a worthwhile conversation. We are working on adjacent problems from different directions, and it’s fascinating to explore how we are all riding these waves of innovation and change together.

CalcTree sits somewhere between Notion and Mathcad. Pages are markdown, calculations run on MathJS with native unit arithmetic, and there is a Python engine behind the page for anything the calculation library does not cover. Multiple people can work the same page live. The AI layer builds calculations, reviews them, and reads from a vectorized workspace knowledge base so it produces work in the context of your own standards and checklists.

Before the call I sent Tim two unlabeled datasets and a Python script with no context. In the episode he turns that into a run-of-river hydro report on the platform, pretty cool.

Other ground covered:

* Adopting a tool like this and making decisions with so many options

* Model selection inside the platform, and how effort and thinking settings trade against a monthly token allowance

* The billable hours problem. If the work takes less time, what happens to the fee.

* Linked references, so a reviewer can click a code clause in a calculation and land on the actual table with its surrounding context rather than a screenshot

* The dependency graph view, and what it’s for

* Quality assurance and control.

I recommend subscribers take a look for themselves, CalcTree is a powerful platform and full of useful features for team-based engineering work. There is a generous free tier with access to the built-in template library, which contains around a hundred calculations across Australian, US and European codes.

Tim like the GLM models, GLM 5.3 has recently released and I think it’s a great model too. Worth checking out on Fireworks.

Great to speak with Tim, as always.

See you in the next one.James 🌊



Get full access to Flocode: Engineering Insights 🌊 at flocode.substack.com/subscribe

Podden och tillhörande omslagsbild på den här sidan tillhör Exploring Python's Power in Engineering with James O'Reilly. Innehållet i podden är skapat av Exploring Python's Power in Engineering with James O'Reilly och inte av, eller tillsammans med, Poddtoppen.