GSoC 2026 — Interest in Scalable Online Bayesian State Space Models | Sujan Kumar MV

[CR1TIC4L]

Hi PyMC community!

I’m Sujan Kumar MV (@KRYSTALM7), Backend Engineer at TCS, M.Tech CS from VIT Vellore. I’m interested in the Scalable Online Bayesian State Space Models project for GSoC 2026.

My research background is directly relevant — I’ve published work on hybrid RL + ML frameworks for time-series state estimation problems in Nature Portfolio (Q1), Springer, and IET.

I’ve read the contributing guide, am setting up PyMC locally, and plan to open my first PR this week.

Looking forward to any guidance from mentors and the community!

Portfolio: https://sujan-space.vercel.app

[CR1TIC4L]

Following up on my introduction — wanted to share more context
on my background and fit for this project.

Relevant work:

• Published research on hybrid RL + ML framework for sequential
  state estimation in EV batteries (Nature Portfolio, 2026):
  Optimizing charge discharge cycles using QPPONet-enabled hybrid learning framework for energy management and safety in electric vehicles | Scientific Reports
• Hybrid LSTM-GRU for time-series prediction (Springer, 2025)
• First PyMC contribution — numpydoc docstring improvements:
  doc: add numpydoc-compliant docstrings to HalfFlat distribution by KRYSTALM7 · Pull Request #8208 · pymc-devs/pymc · GitHub

Why this project: The online state space problem — updating
posteriors as new observations arrive without full re-sampling —
is exactly what my EV battery BMS research solves in a different
domain. I understand this problem deeply from both a research and
implementation perspective.

Time commitment: 350 hours — 2hrs/day weekdays, 7hrs/day
weekends. No exams or conflicts May–September 2026.

Contact:

• GitHub: KRYSTALM7 (Sujan Kumar MV) · GitHub
• Portfolio: https://sujan-space.vercel.app
• Email: sujankumar7702@gmail.com
• LinkedIn: https://linkedin.com/in/sujankumar2003

All application materials: KRYSTALM7 (Sujan Kumar MV) · GitHub