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
Variational Autoencoder for Synthetic Tabular Banking Data
You receive a 500K-row anonymized transaction dataset with 25 columns (mixed numerical + categorical). Train a VAE (TabVAE or a small custom model) with appropriate likelihoods …
- Variational Inference
- Deep Generative Models
- Synthetic Data
Probabilistic Machine Learning - AnalysisBeginnerNew
Explain a Credit-Risk Model with SHAP for a Fintech
You receive a trained XGBoost credit-risk model (binary default prediction), the training feature schema (38 features), and a held-out 10,000-sample test set with labels. Comput…
- Shap
- Interpretability
- Fairness Analysis
Explainable and Interpretable AI - CodeIntermediateNew
Forecast Intraday FX Volatility for a London Liquidity Desk
You receive 18 months of tick-level mid-quote data for six FX pairs plus a calendar of scheduled macro events. Resample to 1-minute bars, engineer realized-volatility features, …
- Time Series Forecasting
- Feature Engineering
- Model Validation
AI and Quantitative Finance - AnalysisSeniorNew
Write a Copyright Risk Memo for a Foundation-Model Lab's Training Set
Cover (1) US fair-use exposure for training on web-scraped text and code, including the current state of pending major lawsuits at the time of writing; (2) the EU TDM exceptions…
- Copyright Law
- Regulatory Analysis
- Risk Mapping
AI Law, Policy, and Regulation Practice your coursework on real scenarios.
Every challenge is shaped from real-world context — not generic exercises. The work mirrors what your degree prepares you for.
Why Ewance
- AnalysisBeginnerNew
Audit a Climate-Tech Sensor Dataset for Production Readiness
You receive 18 months of raw sensor readings from 1,200 sensors (about 800M rows), plus a sensor-metadata table (location, firmware version, deployment date). Profile the data f…
- Data Quality Audit
- Data Profiling
- Time Series Analysis
Applied Data Analysis and Practical Data Science - ResearchSeniorNew
Embodied Visual Reasoning for a Warehouse Pick Assistant
Use an embodied simulator (Habitat 3.0 or Isaac Sim — pick one and justify) to render 300 cluttered-bin scenarios with a target item label. For each scenario, build two reasonin…
- Embodied Vision
- Vision Language Models
- Visual Reasoning
Visual Intelligence and Visual Reasoning - AnalysisBeginnerNew
Cluster a Telco's Subscriber Base for a Pricing Refresh
You receive 12 months of anonymized subscriber-level data: monthly minutes, SMS, mobile data, top-up frequency, top-up amount, churn flag, and tenure. Clean and feature-engineer…
- Clustering
- Feature Engineering
- Exploratory Data Analysis
Data Mining and Knowledge Discovery - DesignIntermediateNew
Design a Real-Time Order Pipeline for a Fintech Payments Platform
You receive a synthetic Kafka stream of around 500 transactions per second, a static merchant dimension table (about 80,000 rows), and a daily FX rate snapshot. Design an end-to…
- Streaming Data
- Kafka
- Stream Processing
Data Engineering and Big Data Systems - Browse challenges
Explore role
Marketing Analyst
Plan and measure campaigns that grow the business. Funnel analytics, attribution, segmentation, and the rigorous measurement that lets marketing defend its budget at the leadership table.
- ResearchSeniorNew
Open-Vocabulary Segmentation Benchmark for a Robotics R&D Lab
Use a curated 200-image household scene set (publicly-available HM3D renderings or COCO + a handful of household prompts). Benchmark 3 open-vocabulary segmentation models: SAM +…
- Open Vocabulary Segmentation
- Vision Language Models
- Benchmarking
Computer Vision - CodeBeginnerNew
Quantize a Vision Model for a Smart-Doorbell SoC
You receive a trained FP32 PyTorch person-detector (mAP 0.74 on a 5k validation set) plus a calibration dataset of 500 unlabeled doorbell frames. Convert to ONNX, then apply pos…
- Quantization
- Model Optimization
- Onnx
Edge ML and On-Device Machine Learning - ResearchIntermediateNew
Audit Recommender Filter Bubbles for a Civic Forum
You receive 90 days of impression logs (about 30 million recommendation events) tagged with content viewpoint labels (left-leaning, center, right-leaning, non-political) from an…
- Recommender Evaluation
- Diversity Metrics
- Audit Methodology
Social Network Analysis and Web Science - CodeBeginnerNew
Structured-Output Prompts for Invoice Extraction
You receive 300 real invoice transcripts (already OCR-ed) labeled with 14 target fields, plus the current production prompt and its 12 percent failure log. Design a new prompt u…
- Structured Output
- Json Schema
- Few Shot Prompting
Prompt Engineering Build a verifiable portfolio.
Submissions become evidence. Reviewers with shipping experience score against a rubric; the result becomes a credential anyone can verify.
Why Ewance
- CodeIntermediateNew
Design an SAT-Based Verifier for an Autonomous-Vehicle Test Lab
Model a simplified four-way intersection: agent positions, lights, and discrete time steps. Define 5 safety properties in propositional logic (e.g., 'no two agents in the inters…
- Sat Solving
- Logical Inference
- Formal Verification
Artificial Intelligence: Principles and Techniques - PresentationSeniorNew
Run a Post-Mortem on a Failed ML Deployment
You receive a packet: original training data sample, post-launch production logs, three Slack-style threads from the on-call rotation, and a summary of the telco's complaints. R…
- Root Cause Analysis
- Stakeholder Framing
- Model Monitoring
Machine Learning in Practice - CodeIntermediateNew
Fine-Tune a 3B Open-Weight Model for Customer Support Triage
You receive 40,000 anonymized labelled support tickets across 18 categories. Fine-tune a 3B open-weight model using parameter-efficient fine-tuning (LoRA) for the classification…
- Lora Fine Tuning
- Open Weight Llms
- Classification
Large Language Models - DesignSeniorNew
Build an Edge MLOps Pipeline for a Smart-Agriculture Sensor
You receive a fleet simulator (1,000 simulated sensors with bandwidth + battery profiles), a model registry stub, and the current firmware's model-loading interface. Design and …
- Edge Mlops
- Ota Updates
- Model Versioning
Edge ML and On-Device Machine Learning - DesignBeginnerNew
Conversational UI for a Personal-Finance Assistant
You will work from 4 scripted scenarios: 'how much did I spend on coffee last month', 'why did my rent payment fail', 'help me set up an emergency fund', and an out-of-scope 'is…
- Conversational Ui
- Dialogue Design
- Trust Design
Question Answering and Conversational Systems - ResearchSeniorNew
Inductive Logic Programming for a Fraud-Rule Discovery Pilot
You receive a labeled fraud dataset (around 25,000 transactions, around 4% positive class), a feature schema (28 features including device, geo, behavioral history), and a basel…
- Inductive Logic Programming
- Symbolic Ai
- Rule Learning
Fuzzy Logic, Knowledge Representation, and Symbolic Reasoning - CodeIntermediateNew
Build a Domain Instruction-Tuning Recipe for a Legal Coach
You will source instruction data from three streams: ~3,000 synthetic paralegal Q&A generated by a frontier model (anonymized prompts), ~1,500 curated examples from public legal…
- Instruction Tuning
- Lora Fine Tuning
- Data Curation
Large Language Models - ResearchSeniorNew
Train Cooperative Agents with Multi-Agent RL
Pick an open multi-agent environment (PettingZoo's MPE 'simple_spread', Overcooked-AI, or SMAC). Implement or wrap three methods: IPPO (independent PPO per agent), MAPPO (centra…
- Multi Agent Reinforcement Learning
- Ppo
- Pytorch
Multi-Agent Systems - CodeIntermediateNew
Agentic RAG with Context-Window Budgeting
You receive a synthetic dataset of 60 founder-style queries paired with 'workspaces' (each up to 500 documents across 3 source types), plus gold-standard answers and citation li…
- Agentic RAG
- Context Window Management
- Iterative Retrieval
Retrieval-Augmented Generation - CodeIntermediateNew
Plan Inventory Replenishment as an MDP for an E-Commerce AI Startup
You receive 18 months of daily demand for 50 representative SKUs at one warehouse plus lead-time and unit-cost data. For one SKU at a time, formulate an MDP with state = (on-han…
- Mdp Modeling
- Value Iteration
- Dynamic Programming
Decision Making Under Uncertainty - AnalysisIntermediateNew
Evaluate an Agent Suite on the SWE-Bench-Style Coding Benchmark
You receive a sandboxed set of 50 small repo-modification tasks (test-passing as the success signal). Run 3 open-source agent frameworks (e.g., OpenHands, SWE-agent, and Aider) …
- LLM Agents
- Agent Evaluation
- Benchmarking
AI Agents and LLM-Based Agents - CodeSeniorNew
Train a GAN for Synthetic Defect Augmentation on a Factory Line
You receive a labeled defect dataset (12 defect types, ranging from 8 to 4,200 examples each), the production classifier, and a starter StyleGAN2-ADA codebase. Train a GAN per r…
- Gans
- Stylegan
- Data Augmentation
Generative AI
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.