Skip to content
NewmathSteps is now on PyPI

Building tomorrow's stack, one useful tool at a time.

Stacktern is an independent technology studio working at the intersection of AI, data and developer tooling — shipping open, practical products in public.

Explore productsFollow on GitHub
Products live
03
In the pipeline
02
Built in public
100%

Products

What we're building

From open models to learning tools and developer libraries — everything shipped under Stacktern, newest first.

Open Source · Python Library

mathSteps

New

Every step explained. Every answer verified.

A step-by-step symbolic and numerical math solver for Python. Linear algebra, root finding, interpolation, integration, ODEs and BVPs — it shows every intermediate step and checks each answer against NumPy, SciPy and SymPy. Use it as a CLI or a library.

  • 21 solvers across 4 domains
  • Exact answers via SymPy
  • CLI + typed Python API
  • MIT licensed
Explore mathStepsView on GitHub
solve.py
import mathsteps

r = mathsteps.linear_system([[1, 2], [3, 4]], [5, 11])

# every intermediate step
for step in r.steps:
    print(step.description)

r.exact      # [1, 2]
r.verified   # True ✓ checked against NumPy
  • Live

    EdTech · AI Tutor

    LearnAI Tutor

    An interactive web app that teaches Data Science and Machine Learning fundamentals through 10 hands-on lessons, live simulators, and a Gemini-powered AI tutor on every page.

    Open live appLearn more
  • Live

    AI · Language Model

    Personal Assistant LLM

    gemma3-270m-personal-assistant-v1

    A fine-tuned Gemma 3 270M language model optimised for personal assistant tasks. Lightweight, open, and ready to run from HuggingFace.

    View on HuggingFace
  • Coming Soon

    Monitoring · SaaS

    Statly

    A dead-simple status page builder that checks your site or API every minute and alerts subscribers when you go down — a professional “All Systems Operational” page for indie hackers and small agencies at $9/mo.

    In development — stay tuned.
  • Coming Soon

    Developer Tool · Data

    Trackline

    Data pipeline observability from a single Python decorator: schema changes, NULL spikes, volume anomalies and lineage, with zero configuration and an optional health dashboard.

    In development — stay tuned.

Approach

How we build

Stacktern is early-stage and independent. A few principles keep the work honest.

  • Open by default

    Models on HuggingFace, libraries on PyPI, source on GitHub. If we build it, you can read it, run it and improve it.

  • Correct before clever

    We test against trusted references and say plainly what a tool does not do. Useful beats impressive.

  • Small, sharp tools

    Each product solves one problem well and fits into the stack you already have, rather than replacing it.

Contact

Have an idea, a bug, or a collaboration in mind?

We read every message. Reach out about our products, partnerships, or anything you'd like to build together.

debojitbasak102@gmail.com
Send an email