December 9–10, 2023 · Cornell Tech, NYC

LlamaHack

The infinite possibility of LLMs is sparking our imagination to rethink every aspect of our lives. LlamaHack was 24 fast-paced hours for students to meet new peers, ship code, and learn new frameworks — while having a bunch of fun.

633
on the guest list
24h
of straight building
$143,596
in prizes
9
winning teams
The LlamaHack llama
Hugging Face
AWS
Cal.com
Supabase

What it was

A weekend of shipping with LLMs

LlamaHack was an immersive event for students exploring cutting-edge large language models — organized by the Cornell LLM Engineering Club on Roosevelt Island.

In December 2023, GPT-4 was a year old, Llama 2 was three months old, and nobody quite knew what an LLM app was supposed to look like yet. So 633 people put their names down, a few hundred showed up on a cold Saturday morning on Roosevelt Island, and 24 hours later there were working prototypes for Medicaid paperwork, news bias, physical therapy, nutrition and calmer arguments.

Nine teams took home a prize. The rest took home a weekend of building with people they had not met on Friday — which was always the actual point.

When
Dec 9–10, 2023 · 24 hours
Where
Cornell Tech, Roosevelt Island, NYC
Who
Students, mostly from Cornell Tech, Columbia and NYU
Deliverable
A pitch and a working prototype, on Devpost by 3:14 PM Sunday

Made possible by

Sponsors

Seventeen teams put up the venue, the credits, the hardware, the mentors and the boba.

How the weekend ran

Schedule

Doors at 8:30 on Saturday, git push deadline at 3:14 PM on Sunday. Pi o’clock, obviously.

Saturday

December 9
  1. 8:30 AM

    Networking + breakfast

  2. 9:50 AM

    Opening speeches

  3. 10:45 AM

    Hack time

  4. 12:00 PM

    Lunch

  5. 1:00 PM

    More hacking + mentor support

  6. 7:00 PM

    Dinner

  7. 7:30 PM

    Continue hacking

Sunday

December 10
  1. 9:00 AM

    Breakfast

  2. 10:00 AM

    Keep pushing

  3. 12:00 PM

    Lunch

  4. 3:14 PM

    Git push deadline

  5. 4:00 PM

    Hackfair — everyone demos

  6. 5:30 PM

    Final demos

  7. 6:30 PM

    Awards & closing ceremony

Twelve ways to win

Prizes

Three podium places plus nine sponsor tracks, so a team could win on craft, on impact, or on sheer taste in tooling.

🥇

1st place

  • $10,000 AWS credits
  • $800 Cohere credits
  • $400 Hugging Face credits
  • 4× Cornell bottles + notebooks
🥈

2nd place

  • $10,000 AWS credits
  • $800 Cohere credits
  • $400 Hugging Face credits
  • 4× Cornell bottles
🥉

3rd place

  • $10,000 AWS credits
  • $800 Cohere credits
  • $400 Hugging Face credits

People’s choice

50 free bobas from Boba Whale · $10,000 AWS credits

Best LLM solution

1× iPad Pro · $10k CloudCodeAI credits · $10,000 AWS credits

Best use of Together AI

4× AirPods Max · $2,500 Together AI credits · $10,000 AWS credits

Best healthcare solution

2× AirPods Pro (2nd gen) · $10,000 AWS credits

Best fork of Cal.com

1× MacBook Air M2 · $10,000 AWS credits

Supabase prize

4× $250 in Supabase credits · $10,000 AWS credits

Best developer tool by Warp

4× Keychron keyboards · 4× Warp mugs · $10,000 AWS credits

Best use of LastMile AI

$200 Amazon gift card · $10,000 AWS credits

Best use of MongoDB Atlas

$10,000 AWS credits

Where the money went

$143,596 in prizes

Almost all of it in cloud and model credits, spread across twelve tracks so more than one team could win something.

Credits

$137,100

  • AWS$10,000 of cloud credits on each of the 12 tracks$120,000
  • CloudCode.aicredits on the Best LLM solution track$10,000
  • Together AIcredits on the Best use of Together AI track$2,500
  • Cohere$800 of credits on each podium place$2,400
  • Hugging Face$400 of credits on each podium place$1,200
  • Supabase4 × $250 of credits on the Supabase prize$1,000

Hardware & swag

$6,496

  • 1× MacBook Air M2
  • 1× iPad Pro
  • 4× AirPods Max
  • 2× AirPods Pro
  • 4× Keychron keyboards
  • $200 Amazon gift card
  • 50 bobas from Boba Whale
  • Cornell bottles, notebooks & Warp mugs

Credit figures are the per-track amounts published on Devpost, summed across the tracks that offered them.

Who picked the winners

Judges

Engineers, researchers and investors who sat through every demo on Sunday evening.

  • Greg Morrisett

    Greg Morrisett

    Dean and Vice Provost at Cornell Tech

  • Daniel Doubrovkine

    Daniel Doubrovkine

    Principal engineer at AWS, former CTO at Artsy

  • Daniel Lobato

    Daniel Lobato

    ML engineer at Meta, former principal engineer at Red Hat

  • Advait Maybhate

    Advait Maybhate

    Software engineer at Warp and Hack the North organizer

  • Alex Reinhart

    Alex Reinhart

    Investor at Pillar VC, former co-founder of Parambil

  • Antoni Viros

    Antoni Viros

    Research scientist at IBM and Cornell alum

  • Saurav Panda

    Saurav Panda

    Co-founder and CEO of CloudCode.ai

  • Zoe Weil

    Zoe Weil

    Applied staff scientist at Etsy, former chief scientist at Risalto Health

Judging criteria

Approach & idea
How viable is the solution and how well does it address the problem? Market strategy, feasibility, creativity.
Product — technical
Is there a demo? How well was the idea prototyped, and how technically impressive is it?
Presentation
Clear and appealing pitches are essential.
Product — design
How much effort and thought went into the design?
X factor
What makes the project extraordinary.