Build Your Own
Coding Agent
The AI Coding Gap
The AI Coding Gap
The AI Coding Gap
On 18 real software engineering tasks, Claude 3.7 Sonnet writes code that passes test cases but is not mergeable.
On 18 real software engineering tasks, Claude 3.7 Sonnet writes code that passes test cases but is not mergeable.
On 18 real software engineering tasks, Claude 3.7 Sonnet writes code that passes test cases but is not mergeable.
The truth? Current AI models weren't built to be autonomous software engineers. They can't efficiently navigate complex codebases, debug multi-step problems, or maintain context across large-scale refactoring tasks.
The truth? Current AI models weren't built to be autonomous software engineers. They can't efficiently navigate complex codebases, debug multi-step problems, or maintain context across large-scale refactoring tasks.
The truth? Current AI models weren't built to be autonomous software engineers. They can't efficiently navigate complex codebases, debug multi-step problems, or maintain context across large-scale refactoring tasks.
39%
39%
39%
Passes all test cases
0%
0%
0%
Mergeable as-is
This is your opportunity to build a coding agent
designed for the real-world.
Your Opportunity
Build an autonomous coding agent that
understands requirements
plans the approach
writes code
identifies errors
fixes errors
runs and validates code
Build an autonomus coding agent that
understands requirements
understands requirements
Build an autonomus coding agent that
understands requirements
understands requirements
The Leaderboard
2
Karthik Bhat
India
🥈
1
Shuhan Cai
USA
🏆
3
Bhavini Goel
India
🥉
4
Ibrahim Kirabo
Kenya
5
Manmohan Negi
USA
6
Duong Bao Duy
Vietnam
7
Srijan Sarmah
India
8
Chetan Singh Dhanjal
India
9
Roshni Kumari
India
10
Syed Farazuddin Hussaini
USA
The Leaderboard
4
Ibrahim Kirabo
Kenya
5
Manmohan Negi
USA
6
Duong Bao Duy
Vietnam
7
Srijan Sarmah
India
8
Chetan Singh Dhanjal
India
9
Roshni Kumari
India
10
Syed Farazuddin Hussaini
USA
🥈
Karthik Bhat
India
2
🏆
Shuhan Cai
USA
1
🥉
Bhavini Goel
India
3
The Timeline
Registration opens
Registration closes
September 4
Invitations sent to selected participants
72-hour coding challenge
Results and prizes announced
The Timeline
Registration opens
Registration closes
September 4
Invitations sent to selected participants
72-hour coding challenge
Results and prizes announced
August 25, 12:00 AM UTC
Registration opens
September 3, 12:00 AM UTC
Registration closes
The Challenge Overview
The Challenge Overview
The Challenge Overview
Selected candidates receive an email invitation to join the challenge.
Selected candidates receive an email invitation to join the challenge.
Selected candidates receive an email invitation to join the challenge.
Build your coding agent using your preferred AI tools/IDE or on the HackerRank online IDE.
Build your coding agent using your preferred AI tools/IDE or on the HackerRank online IDE.
Build your coding agent using your preferred AI tools/IDE or on the HackerRank online IDE.
Push your code to the HackerRank platform within 72 hours of starting.
Push your code to the HackerRank platform within 72 hours of starting.
Push your code to the HackerRank platform within 72 hours of starting.
Evaluation Criteria
Test Cases Passed
Test Cases Passed
Test Cases Passed
We will measure how many test cases your agent successfully completes across different complexity levels.
We will measure how many test cases your agent successfully completes across different complexity levels.
We will measure how many test cases your agent successfully completes across different complexity levels.
Cost Efficiency
Cost Efficiency
Token Efficiency
We will measure how efficiently your agent uses resources. Agents that deliver better results at lower cost will be rewarded.
We will measure how efficiently your agent uses resources. Agents that deliver better results at lower cost will be rewarded.
We will measure how efficiently your agent uses resources. Agents that deliver better results at lower cost will be rewarded.
Who Can Register
Whether you're a Builder or Founder, a ML/AI Engineer, or a Back-End, Front-End, or Full-Stack Developer, this hackathon is for you! Even if you’re just curious about AI, you’re welcome to join in.
This is an
invite-only
event with limited spots available. Selected participants
will be notified on
August 29th.
Whether you're a Builder or Founder, a ML/AI Engineer, or a Back-End, Front-End, or Full-Stack Developer, this hackathon is for you! Even if you’re just curious about AI, you’re welcome to join in.
This is an
invite-only
event with limited spots available. Selected participants
will be notified on
August 29th.
Whether you're a Builder or Founder, a ML/AI Engineer, or a Back-End, Front-End, or Full-Stack Developer, this hackathon is for you! Even if you’re just curious about AI, you’re welcome to join in.
This is an
invite-only
event with limited spots available. Selected participants
will be notified on
August 29th.
The Rewards
Top 3 Winners
-
Gift coupons worth $500 to purchase any AI tool of your choice
-
Early access to HackerRank products
-
Social media shoutout
-
HackerRank Goodies
Results will be announced within 10 days after the event ends.
The Rewards
The Rewards
Top 3 Winners
Top 3 Winners
-
Gift coupons worth $500 to purchase any AI tool of your choice
-
Early access to HackerRank products
-
Social media shoutout
-
HackerRank goodies
-
Gift coupons worth $500 to purchase any AI tool of your choice
-
Early access to HackerRank products
-
Social media shoutout
-
HackerRank goodies
All Participants
All Participants
10 HackerRank AI Mock Interview credits
10 HackerRank AI Mock Interview credits
Results will be announced within 10 days after the event ends.
Results will be announced within 10 days after the event ends.
Frequently Asked Questions
Frequently Asked Questions
Frequently Asked Questions
Is it solo or team-based?
Is it solo or team-based?
Is it solo or team-based?
Do I need prior AI experience?
Do I need prior AI experience?
Do I need prior AI experience?
What if I encounter technical issues during the challenge?
What if I encounter technical issues during the challenge?
What if I encounter technical issues during the challenge?
Will I get API access?
Will I get API access?
Will I get API access?
Can I use my own libraries?
Can I use my own libraries?
Can I use my own libraries?
Can I use my own AI IDE or AI tools for this challenge?
Can I use my own AI IDE or AI tools for this challenge?
Can I use my own AI IDE or AI tools for this challenge?
Register
Register