BangPypers x MariaDB Python Hackathon – Winners Announced!
Last Saturday marked an exciting milestone: the announcement of the winners in our first large-scale MariaDB Python Hackathon, organised in collaboration with BangPypers, HackerEarth, and MariaDB plc. Over the past months, developers from across India have explored new ways to make MariaDB easier to use, more connected, and better integrated into today’s most important open-source ecosystems.
This post celebrates the outstanding contributions in both the Integration Track — projects that help MariaDB work seamlessly with other tools and frameworks – and the Innovation Track – projects that showcase existing MariaDB features and make it easy to learn, copy, and adapt.
Why we organised the MariaDB Python Hackathon
Large organisations often run hackathons to stimulate innovation and uncover fresh ideas that might not emerge through ordinary workflows. They also serve to engage developers, researchers, and partners, fostering a sense of community around the organisation’s technology.
In addition, hackathons can reveal new talent, highlight promising prototypes, and accelerate experimentation — all within a short, energised timeframe. For many companies, they are as much about inspiration and collaboration as they are about the specific solutions produced.
In our case at the MariaDB Foundation, the MariaDB Python Hackathon, held in collaboration with BangPypers, was primarily about making MariaDB even easier to use. Within our Integration Track, the goal was to ensure that MariaDB can be used seamlessly across a wide range of environments, tools, and open source projects. And in the Innovation Track, the goal was to ensure there are role models and reference cases for application developers to copy and adapt.
We got great ideas and surprisingly mature implementations
Robert Silén and I are genuinely delighted with the outcomes of this, our first major Hackathon in India — both in terms of the ideas proposed and the execution achieved. The participants’ creativity has revealed much about what developers want: which cool tools and frameworks do people wish to use MariaDB from? This insight helps us set our priorities and better understand where and how MariaDB is being used in practice.
Compared to our Python Hackathon in Finland, the numbers were huge:
4,189 registered individuals on HackerEarth’s platform led to 597 idea submissions (29% integration, 71% innovation). We short-listed 148 ideas for development, that were best aligned to the hackathon’s goals. At the end of the development phase we received 59 final code submissions (35 innovations, 24 integrations). Evaluating ideas and final submissions was tough out of such a great variety of ideas and final submissions.
The Integration Track winners
1. Our Overall Winner – Airflow Integration by Ctrl_Alt_DB
Pratush Maheshwari (Noida, Uttar Pradesh) and Jyothi Muthuraj (Bengaluru, Karnataka) built the first native MariaDB connector for Apache Airflow. Until now, those wishing to use MariaDB from within Airflow have typically relied on the MySQL connector — but that solution has come with a number of shortcomings. It contains bugs, lacks performance, and does not recognise MariaDB’s unique features. Ctrl_Alt_DB, on the other hand, does — it even supports MariaDB ColumnStore. See Pratush’s and Jyothi’s Pull Request here.
2. Second Place – Dagster Integration by Ria Kulkarni and Revanth Sreeram (Bengaluru, Karnataka)
Ria Kulkarni’s and Revanth Sreeram’s Dagster–MariaDB integration connects MariaDB to today’s perhaps fastest-growing data-orchestration framework. Beyond Airflow, there is now also Dagster to help engineers schedule, monitor, and automate pipelines — making MariaDB part of the modern DataOps stack. See Ria’s and Revanth’s Pull Request here.
3. Third Place – Ollama + MariaDB MCP by Pratik Davidson Deogam (Bhubaneswar, Odisha)
Pratik Davidson Deogam integrated Ollama, the open-source LLM framework, with MariaDB’s MCP Server. It’s a glimpse into how AI workflows can soon talk directly to MariaDB. See David’s Pull Request here.
4. Runner-up: MariaDB Vector integration into Django by Apurv Chaudhary (Uttarakhand)
As runner-up, we are pleased to recognise a project by Apurv Chaudhary from Uttarakhand in northern India: Django-MariaDB-Enhanced-Integration (mariax). Apurv introduces MariaDB Vector support into the highly popular Django framework. Django boasts more than 85,000 GitHub stars — an impressive level of adoption. Apurv doesn’t yet have a Pull Request, so his implementation is in his GitHub repository.
The Innovation Track winners
1. First place – Query optimization for MariaDB Vector by Aakanksha Singh and Mihir Phalke (Mumbai, Maharashtra)
Aakanksha’s and Mihir’s submission addresses a critical question: when should you use Vector search versus traditional SQL queries? They both analyse the problem and suggest an automated solution delivering 1.5x to 16x speedups in a very neat and professional submission. The project directly implements MDEV-33412, a documented feature request for “cost-based optimizer choice for k-NN indexes.” See their impressive submission in their repo.
2. Second place – Multimodal Embeddings with MariaDB Vector by Sunny Kumar and his team (Pune, Maharashtra)
MariaDB Vector is MariaDB’s fastest adapted feature that makes AI and semantic search applications possible. Sunny Kumar’s team explored in their submission Metadata-Hub how to do multimodal semantic search in a professional looking use interface using OpenAI’s CLIP that creates vectors from both text and images. See their repo here.
3. Third place – Temporal Tables with FileVault by Avishkar Patil (Kolhapur, Maharashtra)
Temporal Tables is a MariaDB feature used for automatic versioning and auditing of database changes without application-level tracking code. FlightVault by Avishkar Patil shows how to use Temporal Tables by implementing a visual disaster recovery system that queries historical states of the database.
4. Runner-up – JSON and Window functions by Divyani Singh (Agra, Uttar Pradesh)
Divyani Singh’s showcasing of JSON and Window Functions shows how to optimize performance with JSON indexing with suitable implementations for both MariaDB and MySQL. MySQL-to-MariaDB migration is generally more or less trivial, though some implementation differences exist – so we are happy Divyani highlights this with her submission. See her Github repo.
There is more to come
There is much more to be said about the hackathon, and trust us: There will be interviews, more blogs, and overall analyses. We will share several of the other contestants, as not everyone could be a winner. But for now, we wanted to share the joy and celebrate the winners.
Final Reflections and Thanks
We extend our heartfelt thanks to all participants, mentors, and partners who made this hackathon possible — and in particular to our friends at BangPypers, HackerEarth.and MariaDB plc.
At BangPypers, our sincere appreciation goes to Jayita, Indivar, and Abhilash — for their collaboration, energy, and local engagement, which brought the event to life. And to HackerEarth – Aftab, Tripti, Prajwal, and Abhijit – we owe special thanks for helping us organise a hackathon of this remarkable scale — something we could never have achieved on our own.
We want to thank MariaDB plc, in particular Conor, for believing in us enough to sponsor the event, the organisation with HackerEarth, and the prizes. This made the whole hackathon possible.
Robert Silén and I also wish to thank the specialists from both MariaDB plc and MariaDB Foundation, whose insights were invaluable. Without their expertise, it would have been impossible to find the needle in the haystack among so many worthy contributions. In particular, we’d like to recognise Daniel, Ian, Sergei, Vikas, Adam, Nikita, Hartmut, and Roman for their thoughtful guidance and active involvement throughout the evaluation process.
And most of all: thank you to all the participants in the hackathon! Your work has been truly invaluable. We will be reaching out to several other participants — including some not mentioned here — to explore how we can continue to build on your ideas and help shine a spotlight on your excellent work. Hackathons like this remind us how vibrant the open source community truly is — and how much innovation can emerge when talented people come together with curiosity and shared purpose. We look forward to continuing this journey, supporting contributors in turning their prototypes into upstream pull requests, and welcoming new participants to our next MariaDB Hackathon.
