Frequently asked questions

Straight answers about who we are and how we work. doom2quake is a human builder collective.

What is D2Q / doom2quake?

doom2quake (D2Q for short) is a human builder collective that wins hackathons and bounties. We track every open hackathon and bounty we can find, read each rulebook in full, and map exactly how to win: the prize stack, the judging rubric, and what took the prize last time. Then we build the concept most likely to win and ship a working demo before the deadline. We are people, not a bot or an automated engine. The site you are reading is our public knowledge graph of that research.

How do you decide which hackathons to enter?

We read each event as a sponsor's showcase and decode what the rubric is really rewarding before we commit. We look at the weighting of the judging criteria, the depth of the prize stack (grand prize versus stackable sponsor prizes), the likely field size, and the eligibility fine print. We only enter events where we can credibly max out the heaviest criterion. Just as importantly, we pass on the wrong events on purpose, so we can put our whole effort into the right one. We wrote up the full method in our guide on how we read a hackathon like a sponsor's showcase.

What have you built?

Our most recent build is Atlas, an autonomous data-ops agent we made for the All Things Agentic Hackathon on Google ADK, Gemini, and BigQuery. Atlas runs a closed operating loop over a data warehouse: it predicts the expected shape of each load, detects anomalies, diagnoses the root cause, quantifies the business impact, remediates what it can do safely, and verifies that the fix held. Atlas itself is a genuinely autonomous agent. We, the team that built it, are humans. There is a full writeup on the blog.

Do you do bug bounties and audit contests?

Yes. Alongside hackathons we track security bug bounties, smart-contract audit contests, and open-source feature bounties, sorted by reward ceiling and scope. We treat them the same way we treat hackathons: read the scope in full, understand what a valid, high-severity finding looks like for that program, and focus where the expected value is highest. Our live bounty board lists the programs we are watching.

How do you approach a hackathon's judging rubric?

We treat the rubric as the actual specification of what wins, weighted. We read the weights rather than just the category names, because the same four categories at different weights are different events. We find the single criterion we can score highest on and build to it, and we watch for criteria that are secretly pass/fail gates, like 'a working demo is required'. If the event has run before, we study past winners as the truest signal of how judges actually scored, then aim one notch above that bar.

Are you an AI, a bot, or an automated system?

No. doom2quake is a human builder collective. We use tools, and we build autonomous agents as products (Atlas is one), but the team making the decisions and shipping the work is people. When we say 'we', we mean the humans in the collective.

Where can I see the hackathons and bounties you track?

The hackathons page is a continuously-updated, filterable board of open hackathons with prizes, judging rubrics, and how to win each one, across every major platform (Devpost, Devfolio, ETHGlobal, DoraHacks, MLH, Unstop, HackerEarth, Kaggle, and Taikai). The bounties page lists live security, audit-contest, and open-source bounties by reward ceiling. Both are linked from the top navigation.