How well can your model understand a doctor’s note? Our benchmark tests how accurately models can identify and link medical concepts using SNOMED CT on the largest public dataset of its kind. Get started: www.drivendata.org/benchmarks/3...

Excited to be speaking at Good Tech Summit in DC April 7 www.goodtechtogether.org/summit We’ll share a program focused on K-12 education and talk about investing in the foundations of AI: data, models, and benchmarks. We'll explore how these shape AI development in a field. Join us!

The performance gap in children’s ASR is real — and solvable! With 1 week until the April 6 deadline, we’re inviting the global ML community to help close it. Compete for $120K and contribute to better speech technology for kids. kidsasr.drivendata.org

Children’s speech remains one of ASR’s toughest challenges — and the leaderboard is moving! 3 weeks left to compete for $120K in the On Top of Pasketti Challenge. Submit your model: kidsasr.drivendata.org

10 years ago, "data science for social good" was just an idea. Today, it's a global movement. Our 10-Year Impact Report reflects on a decade of responsible, real-world AI. Take a look back with us: s3.amazonaws.com/drivendata-p...

Competing in the On Top of Pasketti Word Track? We've published a reference tutorial walking through how to build a model for children's speech recognition — covering data exploration, model training, and submission packaging. Get started: drivendata.co/blog/child-a...

From Bogotá to Almaty, our community continues to impress! IGCPHARMA built prize-winning early dementia prediction models in PREPARE. Kirill Brodt has 23 competitions, 8 top-10s, and 6 top-3 finishes. Two new Community Spotlights: drivendata.co/blog/communi... drivendata.co/blog/communi...

Automatic speech recognition struggles with children’s speech. That gap matters! The $120K Children’s Speech Recognition Challenge is driving progress toward models that truly understand kids. Join breakthrough! Submit your solution by April 6th. kidsasr.drivendata.org

A machine learning competition for NASA sparked something bigger. CyFi (cyanobacteria finder) is now an open-source tool using Sentinel-2 data to monitor harmful algal blooms worldwide, with local validation underway. How we got here: drivendata.co/blog/cyfi-sm...

What happens when AI agents enter ML competitions? Spoiler: Humans 1, Agents 0.25–0.93. The top of the leaderboard — where “good” becomes “great” — still looks very human. How might that change? See the results: drivendata.co/blog/ai-agen...

We’re launching our first benchmark competition. The SNOMED CT Entity Linking Benchmark evaluates how well models structure clinical notes using the SNOMED CT nomenclature for a large, de-identified dataset of annotated records. Help set the baseline: www.drivendata.org/benchmarks/3...

Building data pipelines is deceptively complex; Cold starts, unstable inputs, shifting requirements, and delivery trade-offs create friction at every step. To ease the pain, we examined five recurring challenges and suggest practical improvements: drivendata.co/blog/pipelin...

ML competitions moved fast in 2025 - from 512-GPU training runs to benchmark-style challenges. The new State of Machine Learning Competitions report explores what winning teams used. We're grateful to be part of such a dynamic ML community. Read: mlcontests.com/state-of-mac...

The gap between public data and usable data is the “last-mile data problem.” We’ve all seen it: confusing CSVs, messy schemas, opaque data dictionaries. We’re testing a “baked data” approach to solve it, with promising results. See our recipe: drivendata.co/blog/last-mi...

We gave AI agents 24 hours on the leaderboard. Claude Code (Opus 4.5) and Codex (GPT 5.2) produced dozens of submissions. Some hit the top 20. Others overfit or plateaued. We identified 9 obstacles and 6 open questions. Do they match your experience? drivendata.co/blog/ai-agen...

Voice-based tools have the potential to support learning, accessibility, and early literacy, but only if they work for children. In the $120k On Top of Pasketti Children’s Speech Recognition Challenge, solvers will build ASR systems that understand kids. kidsasr.drivendata.org

Hang out with us at #SciPy2026 this summer! Senior Data Scientist Katie Wetstone is co-chairing the Environmental, Earth, and Climate Sciences track, which you can submit to by February 25.

We're excited to be a part of the K-12 AI Infrastructure Program, advancing open datasets, models, and benchmarks for AI in teaching & learning. The first RFP is now open ($50K–$250K) - apply now! k12-ai-infrastructure.org/rfp-due-marc...

Kids learn through voice, but today's ASR tech can barely understand them. In a new data science challenge, solvers will develop models that work with children's unique speech patterns and compete for a share of the $120k prize pool. kidsasr.drivendata.org

🚨 New opportunity: Help build open-source speech recognition AI 🎙️📚 @drivendata.org is hosting a data science competition to advance automatic speech recognition (ASR) for early education. Two tracks, real impact, and $120K in prizes. Learn more & compete: kidsasr.drivendata.org

DrivenData's Katie Wetstone will be co-chairing the climate sciences track at @scipyconf.bsky.social, where you can share YOUR work in environmental data science!  Submit a #SciPy2026 talk, tutorial, or poster by February 25. See you there!

The $10k prize pool Poverty Prediction Challenge sponsored by The World Bank tackles a major challenge in development research: How do you estimate current poverty rates without recent household expenditure data? Submissions open through midnight UTC February 4, 2026.

In our newest machine learning challenge, solvers will use survey data and help uncover imputation methods for monitoring poverty trends. The top teams will take home a share of the $10,000 prize, provided by The World Bank. Learn more and submit predictions here: www.drivendata.org/competitions...

Throwback to our "Hateful Memes" challenge where teams detected harmful content combining text and images. Critical work for online safety! 🛡️💻 #ContentModeration #AI drivendata.org/competitions/64/

Hot topic in #DataScience: Multi-modal learning is bridging text, images, and audio. Our blog explores practical applications beyond the hype! 🎭🔗 #MultiModal #AI drivendata.co/blog.html

Our latest blog dives into "Causal Inference for Data Scientists" - moving beyond correlation to understand what actually drives outcomes! 🔗💡 #CausalInference #DataScience drivendata.co/blog.html

Ever wonder how AI helps track endangered species? Our "Pri-matrix Factorization" competition used camera trap data to identify primates in the wild! 🐒📷 #WildlifeConservation #ComputerVision drivendata.org/competitions/49/

Community Spotlight: Kirill Brodt The Community Spotlight features fantastic members from our DrivenData community. Kirill Brodt, a researcher in computer graphics at the University of Montreal, talks animation, pose estimation, and data science challenges.

Want to build ML systems that actually work in production? Download our free ebook "The 10 Rules of Reliable Data Science" and learn the essentials from our years of real-world experience. Game-changing insights await! 📊🔬 #DataScience #MachineLearning

Check out our "Open AI Caribbean Challenge" winners! Teams used satellite imagery to map building footprints for disaster preparedness. Amazing work applying #ComputerVision to humanitarian needs! 🛰️🏠 #DataForGood drivendata.org/competitions/58/

📢 We're Hiring: Child Speech Transcriber (Remote, Part-Time) Join DrivenData in making speech technology more accessible for children! We're looking for a Child Speech Orthographic Transcriber to help advance Automatic Speech Recognition (ASR) for young learners. docs.google.com/forms/d/e/1F...

Why do "accurate" models fail in production? Our blog posts dive into better evaluation metrics for real-world ML applications. Essential reading! 📊🧠 #DataScience #MLOps drivendata.co/blog.html

Looking for time series forecasting inspiration? Check out the innovative approaches from our "Power Laws" competition where teams predicted electricity demand across multiple regions! ⚡📈 #DataScience #TimeSeries drivendata.org/competitions/53/

🚀 Stellar news: DrivenData is excited to be selected as one of NASA’s newest open innovation contract awardees! We’re proud to push the boundaries of crowdsourced science & AI innovation to help solve NASA’s toughest challenges & to work alongside this amazing community.

BIG NEWS! 📸 Zamba Cloud now supports camera trap IMAGES! Our wildlife AI recognizes 178 taxa with automatic bounding boxes. Try it at www.zambacloud.com or check our Python library: github.com/drivendataor... #WildlifeConservation #AIforGood

🧠 Zamba Cloud's new image support uses deep learning to detect & classify 178 wildlife taxa. Built on 15M+ annotations from @LILAscience datasets. Customize with transfer learning! Check out our Python library: github.com/drivendataor... #MachineLearning #ComputerVision

Every hour spent manually sorting camera trap images is time NOT spent on conservation action. 🌍 Zamba Cloud now automates wildlife ID for BOTH videos & images, giving researchers their time back. www.zambacloud.com #ConservationTech #BiodiversityProtection 🦏

You don't need to be a coder to use AI for wildlife research! 💻➡️🚫 With Zamba Cloud's new image support, simply upload photos, get species IDs, and even train custom models—all without writing a single line of code. www.zambacloud.com #WildlifeResearch

You don't need to be a coder to use AI for wildlife research! 💻➡️🚫 With Zamba Cloud's new image support, simply upload photos, get species IDs, and even train custom models—all without writing a single line of code. www.zambacloud.com #WildlifeResearch

BEFORE: Spend weeks manually reviewing 10,000 camera trap photos AFTER: Upload to Zamba Cloud, grab coffee, get results ☕ Our wildlife ID system now works for images too! www.zambacloud.com #ResearchEfficiency #WildlifeMonitoring 📊🐾

What's "fine-tuning"? 🧐 It's like teaching Zamba Cloud to recognize YOUR local animals! Upload labeled examples & our AI learns what matters in YOUR ecosystem. Now available for images too! www.zambacloud.com #AIforConservation #MachineLearning 🦁🌿

Calling all #conservation developers! 👩‍💻 Zamba Cloud's image support is powered by our open-source library. Want to customize? Contribute? Build your own tools? We've got you covered: github.com/drivendataor... #OpenSource #ConservationTech 🐯

😍 New DrivenData Labs data services site is LIVE! So much easier to find what we do and see our history of work. Come check it out (and let me know if you see any issues!). drivendata.co

We need LLM benchmarks with the same lessons we learned from ML challenges. Absolutely private test sets, submissions can only be selected based on public score, number of subs limited. Even with these, value decays over time.

AI alone won't change the social sector. I put down a few thoughts spurred at #GTF2025 on what other infrastructure—technical and social—we need to make AI work. Let me know your thoughts! www.linkedin.com/pulse/tech-i...

The "On Cloud N" competition showcased amazing solutions for predicting nitrogen dioxide levels using satellite data. Browse the winning approaches to see how data science tackles air quality monitoring from space! 🛰️☁️ drivendata.org/competitions/142/competition-cloudnine/ #DataForGood

Curious about ML for conservation? Check out the closed "Tick Tick Bloom" competition where data scientists predicted harmful algal blooms in lakes. See how winners combined satellite imagery and water quality data! 🌊🔬 drivendata.org/competitions/143/tick-tick-bloom/ #DataForGood

Tracking elephants in Africa? 🐘 Monitoring wolves in Europe? 🐺 Studying jaguars in the Amazon? 🐆 Zamba Cloud now identifies wildlife in BOTH videos & images across 178 taxa. Try it today: www.zambacloud.com #WildlifeMonitoring #FieldResearch

It took months of work, 15M+ annotations, and countless cups of coffee ☕ to bring image support to Zamba Cloud. Our team is thrilled to help conservation researchers worldwide identify wildlife faster! www.zambacloud.com #ConservationTech #AIforGood 🌍

📢 New #Datachallenge alert: Can you write abstracts for academic social science papers using a large language model that runs on your computer? OK, but can you do it well? 👋 This is a practice competition designed to be accessible to participants at all levels!