Computer & Information Sciences
Data Science Challenges
Real data-science projects and challenges on Ewance — clean messy datasets, build and evaluate models, and turn raw data into decisions the way a working data scientist does. Solve them to build a portfolio of verified, recruiter-checkable proof you can do the work — not just describe it.
Recommended challenges
- CodeIntermediateNew
Lane-Change Intent Classifier from Dashcam Video
Use a public driving video dataset (e.g., Argoverse 2 sensor or BDD100K) and curate ~6,000 short clips labeled with the three-class intent. Train a temporal model (e.g., a small…
- Video Understanding
- Temporal Modeling
- Model Evaluation
Visual Intelligence and Visual Reasoning - AnalysisIntermediateNew
Optimize Stop-Loss Policies with Dynamic Programming at a Quant Fund
You receive five years of daily PnL series for 12 momentum strategies plus a small set of state features (rolling vol, drawdown, regime indicator). Calibrate a discrete Markov m…
- Dynamic Programming
- Backward Induction
- State Modeling
Decision Making Under Uncertainty - AnalysisBeginnerNew
Analyze a Learning-Analytics Dataset for At-Risk Detection
You receive an anonymized dataset of LMS engagement features (logins, assignment submissions, forum posts, video-watch time), grade history, and a binary label for end-of-semest…
- Learning Analytics
- Classification
- Fairness Metrics
AI in Education and Learning Analytics - CodeIntermediateNew
Build a Vector-Search Backend for an Enterprise AI Knowledge Assistant
You receive a corpus of around 20,000 PDFs (mixed scanned and digital) totalling around 30 GB and a labeled retrieval set of 200 queries with human-judged ground-truth passages.…
- RAG
- Vector Search
- Embeddings
Data Engineering and Big Data Systems Develop in-demand professional skills.
Each challenge names the skills it strengthens. Over time, your profile fills with the competences a hiring manager would actually look for.
Why Ewance
- CodeBeginnerNew
Behavior Cloning for a Pick-and-Place Manipulator
You receive 200 human teleoperated demonstrations (state + action trajectories) of picking 8 small electronic components from a tray and placing them at marked locations in a ro…
- Behavior Cloning
- Imitation Learning
- Manipulation
Robot Learning - DesignBeginnerNew
Detect Sensor Drift for a Field Inspection Robot Fleet
You receive 60 days of telemetry from 12 robots, including IMU readings, camera exposure stats, and the inspection-quality scores produced downstream. Define drift signals (roll…
- Anomaly Detection
- Change Point Detection
- Sensor Fusion
Robot Perception and Autonomy - StrategyIntermediateNew
Design a PETs Strategy for an EU AI Act Use Case
Map the underwriting use case to applicable PETs across the data-lifecycle stages (training, evaluation, inference, monitoring). For each, document: privacy property gained, acc…
- Pets Strategy
- Differential Privacy
- Federated Learning
Privacy-Preserving Machine Learning - CodeSeniorNew
PPO Alignment Loop with a Pretrained Reward Model
You receive a small open-weights base model (around 7B), a previously trained reward model, and 5,000 prompts (no responses) for PPO rollouts. Run PPO with TRL's PPOTrainer for …
- Rlhf
- Ppo
- Reward Hacking
Machine Learning from Human Preferences (RLHF and Alignment) - Browse challenges
Explore role
Strategy Analyst
Frame the business question, model the options, build the recommendation. From market sizing to competitive analysis, this role is where strategy consulting meets in-house decision-making.
- PresentationBeginnerNew
Storytelling Visualization of an Autonomous Vehicle Test Campaign
You receive aggregated test results: 12,000 test runs across dry, wet, and snow conditions, with metrics for disengagement rate, near-miss count, and route-completion percentage…
- Data Storytelling
- Audience Adaptation
- Chart Design
Data Visualization - DesignBeginnerNew
Privacy-Preserving Crowd-Density Estimator for Transit Stations
Use a public crowd-counting dataset (e.g., ShanghaiTech or JHU-CROWD) to train a small crowd-density estimator (CSRNet or similar). Wrap it in an on-device pipeline (Python is f…
- Crowd Counting
- Scene Understanding
- Privacy By Design
Visual Intelligence and Visual Reasoning - CodeIntermediateNew
Adapt Machine Translation to a Niche Domain
Pick an open MT base (NLLB-200 or a strong open M2M model). Build a parallel corpus of around 8,000 sentence pairs from the company's bilingual safety standards. Fine-tune on th…
- Machine Translation
- Domain Adaptation
- Transformers
Natural Language Processing - ResearchSeniorNew
Certify Robustness for a Medical-Imaging Classifier
You receive the classifier (a PyTorch ResNet variant) and a 4,000-image labeled validation slice. Apply randomized smoothing (Cohen et al.) at sigma in {0.25, 0.5, 1.0}. Report …
- Certified Robustness
- Randomized Smoothing
- Formal Verification
Trustworthy AI, Robustness, and Safety Get recognized by recruiters and employers.
Credentials are blockchain-anchored via LearnCoin — tamper-evident, portable, link-shareable on LinkedIn and beyond.
Why Ewance
- DesignIntermediateNew
Train a Self-Play Agent for a Card-Game Edtech Demo
Implement a small two-player imperfect-information card game (Kuhn poker or a 3-card simplified Hold'em variant). Implement CFR or CFR+ for the game and run self-play for at lea…
- Counterfactual Regret Minimization
- Self Play
- Game Theory
Artificial Intelligence: Principles and Techniques - CodeIntermediateNew
ReAct Agent for Legal-Research Tool-Use
You receive 30 research questions with paralegal-written gold answers and citation lists, plus stubbed implementations of the 4 tools (you do not need to build retrieval — just …
- React Prompting
- Tool Use
- Agent Design
Prompt Engineering - ResearchIntermediateNew
Multi-Task Learning for a Healthtech Triage Model
You receive 40,000 anonymized de-identified intake-form records with two labels: urgency tier (4 classes) and routed sub-specialty (12 classes). Train (1) two independent classi…
- Multi Task Learning
- Transfer Learning
- Transformer
Meta-Learning, Transfer Learning, and Multi-Task Learning - AnalysisSeniorNew
Brain-Tumor MRI Segmentation Bake-Off
You receive a curated public multi-modal MRI brain-tumor cohort (~600 patients, T1/T1c/T2/FLAIR with whole-tumor / tumor-core / enhancing-tumor masks). Train all three architect…
- Medical Imaging
- Segmentation
- Convolutional Neural Networks
Machine Learning for Imaging and Medical Image Analysis - CodeIntermediateNew
Build a Cross-Lingual Retrieval-Augmented QA System
Index around 5,000 internal-knowledge docs across the three languages using a multilingual embedding model (e.g., multilingual-e5 or BGE-M3). Build the retrieval-then-answer pip…
- RAG
- Cross Lingual Retrieval
- Multilingual Embeddings
Neural Networks for NLP - DesignBeginnerNew
Scaling a Sydney D2C Cosmetics Startup's Data Pipeline
You are tasked with designing a cloud-based data pipeline for GlowUp. The pipeline must ingest real-time user events (page views, purchases, returns) from web and mobile apps, p…
- Cloud Computing
- Apache Spark
- Nosql
Big Data and Cloud Technologies - DesignIntermediateNew
Design a Continuous Eval Pipeline for an Enterprise RAG Product
Design (and partially build) a continuous-eval pipeline for a RAG system: (1) a structured eval set with at least 50 queries grouped by query class; (2) automated scoring (LLM-a…
- Continuous Evaluation
- LLM Evaluation
- Retrieval Augmented Generation
AI Measurement and Evaluation - AnalysisBeginnerNew
Right-Size a Real-Time Recommendation Serving Cluster
You receive 7 days of request-level telemetry (timestamp, latency, error code, pod) plus the existing Horizontal Pod Autoscaler (HPA) and node-group configs. Analyze traffic pat…
- Model Serving
- Kubernetes
- Autoscaling
Machine Learning at Scale - ResearchIntermediateNew
Sim-to-Real Domain Randomization for a Mobile Robot
You receive an Isaac Sim navigation environment, a baseline trained policy, a 50-episode real-bench test set (recorded sensor streams + ground truth) for offline policy evaluati…
- Domain Randomization
- Sim To Real
- Robot Navigation
Robot Learning - CodeIntermediateNew
Fine-Tune a Sequence-to-Sequence Model for Code-Doc Generation
Take a small base model (CodeT5+ or a distilled CodeLlama-Instruct). Build the dataset by mining around 8,000 high-quality function-docstring pairs from permissively-licensed Py…
- Seq2seq
- Transformers
- Lora Fine Tuning
Neural Networks for NLP - DesignBeginnerNew
Design an Automated Essay-Feedback System
You receive 20 anonymized middle-school essays scored by 2 human teachers on a 4-dimension rubric (structure, evidence, voice, mechanics). Design an LLM-based feedback system th…
- Automated Assessment
- Rubric Design
- Prompt Engineering
AI in Education and Learning Analytics - CodeFoundationalNew
Optimizing Inventory for a Barcelona D2C Cosmetics Brand
You are given a CSV file with 6 months of daily sales data for 20 SKUs, including product name, date, units sold, and current stock level. Your task is to write a Python program…
- Python
- Data Cleaning
- Data Analysis
Programming for Business Applications
How it works
From brief to credential, in six steps.
Step 01
Browse challenges aligned to your studies.
Step 02
Accept the one that fits your goals.
Step 03
Work through it with AI Copilot guidance.
Step 04
Submit for structured evaluation.
Step 05
Earn a verified credential.
Step 06
Add it to LinkedIn with one click.
Industry teams behind a decade of practitioner briefs
Hiring from this pool?
Sponsor a challenge and meet candidates through actual work.
Industry teams can shape briefs around the skills they hire for, then evaluate students on rubric-scored deliverables — not resumes.