Technocrats Hackathon

DEFAULT

1st Prize

Domain Sponsor

2nd Prize

3rd Prize

Best Web 3 Website

Best Intelligent AI Build

Best First-Time Builder

Amazon Web Services

**What counts as a strong AWS project?**

- Creative use of AWS services.
- Demonstrating scalability: showing how AWS can take the project from hackathon demo to production.

**Submission Requirements**

- Project must be submitted on Devfolio with **the relevant AWS track selected**.
- Include a short note in your project submission describing which AWS services you used and why.
- Attach the screen recording or screenshots of the AWS architecture, and also include the AWS Builder ID.


**Judging**

- Winners are chosen by Devfolio.
- Criteria: creativity, impact, execution, and effective use of AWS tools.
- Winners announced within **7 days** of hackathon conclusion.

Best AWS Hack

ETHIndia

Select this track if you're building on Ethereum to be eligible for the prizes!

Ethereum Track

Best hack built on Ethereum

EVVM

**EVVM** let's anyone create Virtual Blockchains on top of any chain just deploying smart contracts. For this hackathon we use Ethereum Sepolia.

**EVVM** is a new way to create virtual blockchains on top of existing networks like Ethereum, Arbitrum, Hedera and others. It lets you spin up your own blockchain environment by simply deploying smart contracts instead of running a whole new network. You can have the benefits of a custom blockchain without managing physical infrastructure (no validators or dedicated nodes needed).

Each **EVVM** instance runs as a set of Solidity contracts on a host chain and emulates core components of a blockchain, inheriting the security of the underlying chain. A “blockchain within a blockchain”, similar to a virtual machine in computing, but for blockchains.

**EVVM** is implemented entirely in Solidity, easy for developers to deploy, customize, and upgrade. There’s no need to set up new nodes or consensus mechanisms. This infra‐less deployment approach lowers the barrier to experimenting with new blockchain ideas or tailored networks for specific use cases.

Multiple **EVVM** instances can run in parallel on the same host chain, achieving vertical scalability by stacking virtual chains on one physical network.

Developers can also use **EVVM** as a sandbox to experiment with new features or Ethereum improvements (like testing new opcodes or economic models) in an isolated environment, without affecting the main chain.

Sepolia **EVVM** Contract (will be upgraded for the hackathon):
0xF817e9ad82B4a19F00dA7A248D9e556Ba96e6366

Sepolia MATE Staking Contract (will be upgraded for the hackathon):
0x8eB2525239781e06dBDbd95d83c957C431CF2321

Sepolia NameService Contract (will be upgraded for the hackathon):
0x8038e87dc67D87b31d890FD01E855a8517ebfD24

Faucet for Sepolia MATE tokens: https://evvm.dev

Or Telegram Devs Group https://t.me/EVVMorg

Most Innovative Use of EVVM’s Execution Function - $100

For the project that most creatively integrates **EVVM's** unique transaction execution mechanism, the "executor".
For example, a Fisher/Relayer who executes transactions on the **EVVM** or a new **EVVM** Service that executes on behalf of users under certain circumstances.
**EVVM** can decentralize transaction execution to its network of participants (Fishers/Relayers), Services, or third parties. These should validate user-signed payloads (to prevent reverts) and call **EVVM** contracts to process them when appropriate.
This category invites hackers to be creative with their use of this "executor" model.
We are looking for innovative ideas that enhance the **EVVM** execution flow or extend it in a novel direction.

Workshop:
https://youtube.com/watch?v=uBjvoH5tpJk

Quickstart for Devs
https://evvm.info/docs/QuickStart

MATE Faucet & Tooling
https://evvm.dev

Developers Group in Telegram
https://t.me/EVVMorg

Complete Documentation
https://evvm.info/docs/intro

EVVM Website
https://evvm.org

EVVM Track Execution Function

Most Innovative Use of EVVM’s Execution Function

Best Implementation of Asynchronous Nonces - $50

**EVVM** supports both **synchronous** and **asynchronous** transaction nonces for submitting transactions, and this prize goes to the hack that makes the cleverest use of async nonces specifically.

Asynchronous nonces allow users to issue multiple transactions without waiting for each to complete (out-of-order execution with protection against replay).

This feature can enable parallel transaction processing, user-defined ordering of operations, or optimistic interactions that don’t strictly sequence transactions. Hacks in this category might include a demonstration of payment batching, a game or application logic that relies on out-of-order commits, or tooling that manages async vs. sync nonce flows for developers.
We will favor a project that clearly showcases why asynchronous nonces are powerful. E.g. achieving something that would be cumbersome or impossible with traditional sequential nonces. Implementations should highlight **EVVM’s** flexibility in transaction scheduling and how that can benefit dApp design or performance.

Workshop:
https://youtube.com/watch?v=uBjvoH5tpJk

Quickstart for Devs
https://evvm.info/docs/QuickStart

MATE Faucet & Tooling
https://evvm.dev

Developers Group in Telegram
https://t.me/EVVMorg

Complete Documentation
https://evvm.info/docs/intro

EVVM Website
https://evvm.org

EVVM Track Asynchronous Nonces

Best Implementation of Asynchronous Nonces

Duality

**Duality AI Space Station
Hackathon**
Train cutting-edge AI models for Object Detection in a space station environment using Duality AI's synthetic dataset from their digital twin simulation platform - Falcon.
# Ready to Launch?
Join Duality AI's Falcon platform and start training your space station object detection model with cutting-edge synthetic data!

[Join The Falcon Platform- Start Training ](https://falcon.duality.ai/auth/sign-up?utm_source=hackathon&utm_medium=instructions&utm_campaign=technocrats)


# Challenge Overview
**Space Station AI Training:-**
Duality AI presents a challenge designed to test cutting-edge AI training techniques for Object Detection in a space station environment. Participants will train models using synthetic datasets from Duality AI's digital twin simulation platform, achieving accuracy and precision through parameter optimization.
1.**AI Training**
Learn industry-proven techniques to train AI models using only synthetic data
2.**Digital Twins**
Work with realistic virtual representations of space station environment
3**.Object Detection**
Detect and classify critical space station objects for operational safety

# Challenge Objectives
**Primary Goals**
- Train a robust object detection model using the provided synthetic dataset
- Accurately detect and classify space station objects for operational safety
- Evaluate model performance across challenging scenarios including varied lighting and occlusions
- Benchmark and optimize for accuracy, generalizability, and efficiency

# Hackathon Structure
AI Engineering
• Train and fine-tune YOLOv8 model
• Evaluate model performance
• Optimize accuracy and inference time
• Use synthetic dataset from Falcon
Documentation
• Document complete workflow
• Create structured reports
• Prepare final presentations
• Include performance visualizations
onus Challenge
• Create application using the model
• Plan for model updates with Falcon
• Earn up to 15 bonus points
• Showcase real-world implementation

# Judging Criteria

**Model Performance:-**
**80 Points**
mAP@0.5 (Mean Average Precision at IoU=50%)

Measures accuracy of object detection and classification. Higher mAP scores indicate better performance.

**Report Clarity**
**20 Points**
Structured Findings & Detailed Reporting

Well-organized documentation of methodology, challenges, and solutions with clear steps.

# Key Deliverables

Trained Model
• Fully trained YOLOv8 model
• Model weights and config files
• Training and inference scripts
• Three object category detection

Performance Report
• mAP@0.5 scores
• Confusion matrix analysis
• Failure case evaluation
• Performance visualizations

Documentation
• README with instructions
• Methodology documentation
• Environment requirements
• Reproduction guidelines

# Ready to Launch?
Join Duality AI's Falcon platform and start training your space station object detection model with cutting-edge synthetic data!

[Join The Falcon Platform- Start Training ](https://falcon.duality.ai/auth/sign-up?utm_source=hackathon&utm_medium=instructions&utm_campaign=technocrats)


Duality AI Track

Best Hack Build on Duality AI

Technocrats Hackathon

Technocrats Hackathon