Benchmarking
If you like applying Benchmarking, every challenge here gives you a chance to practice it on a real industry brief.
- CodeSeniorNew
Vectorize a Risk-Calculation Engine for a Mid-Cap Bank
Receive the existing pricer (C++17, scalar, around 600 lines), 2 weeks of representative input data (around 50M paths), the validation harness, and access to an Intel Ice Lake-S…
- Vectorization
- Avx 512
- Simd
Performance Engineering of Software Systems - AnalysisSeniorNew
Spike Three Multiplayer Backend Patterns and Recommend One
Using the gameplay-shape specification, the starter server module, and the simulated player-input traces provided to you, implement minimal but faithful versions of all three ne…
- Realtime Systems
- Architecture Spike
- Networking Patterns
Open coursework - CodeSeniorNew
Diagnose Heap Corruption in a Trading Firm's Order Gateway
Working only from the materials in this case file, reconstruct what is corrupting the heap. Read the representative order-pool allocator module (item: gateway-allocator-module) …
- Cpp Programming
- Manual Memory Management
- Pointers
Open coursework - CodeSeniorNew
HyperLogLog Cardinality for Real-Time Ad-Auction Analytics
Implement HyperLogLog with precision parameter p in {12, 14, 16} (4KB, 16KB, 64KB sketches) and benchmark relative error on a replayed 3-hour production trace (around 13 billion…
- Hyperloglog
- Probabilistic Data Structures
- Streaming Systems
Randomized Algorithms 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
- CodeSeniorNew
Min-Cost Max-Flow for Cloud Workload Placement
Receive 30 days of anonymized placement requests (workload CPU/memory shape, affinity rules), zone capacity per day, and cross-zone bandwidth costs. Model the placement as a min…
- Network Flows
- Algorithm Analysis
- Graph Algorithms
Advanced Algorithms - AnalysisSeniorNew
Amortized-Analysis Investigation of a Production Telemetry Cache
Working only from the three provided materials — the de-identified incident record (incident-record), the representative cache source module (cache-source), and the recorded ope…
- Amortized Analysis
- Data Structures
- Algorithm Analysis
Open coursework - CodeSeniorNew
Port a Numerical Kernel from CPU to GPU for a CFD Simulator
Receive the existing CFD solver (C++17 + OpenMP, around 8,000 lines, the hot kernel is a 7-point stencil sweep over a 512^3 grid), the validation harness, and access to an A100 …
- Gpu Programming
- Cuda
- Parallelism
Open coursework - CodeSeniorNew
Build a Lock-Free MPMC Queue for a Trading-Backend Hot Path
Implement Vyukov's bounded MPMC queue in C++20 using atomic operations with explicit memory orderings (relaxed where safe, acquire/release on the enqueue/dequeue handshake). Use…
- Lock Free Algorithms
- Memory Models
- Cpp
Open coursework - Browse challenges
Explore role
Product Manager
Ship product that solves real user problems. Combine user research, prototyping, and stakeholder alignment to turn ambiguous briefs into measurable wins — the role at the centre of modern software teams.
- AnalysisSeniorNew
Memory Consistency Model Audit of a Lock-Free Queue
Read the SPSC queue source (around 200 lines of C++). For each atomic operation, classify the required ordering and verify the chosen memory_order is sufficient under both x86-T…
- Memory Consistency
- Concurrency
- Lock Free Programming
Advanced Computer Architecture - CodeSeniorNew
Port a CPU Monte-Carlo Simulator to GPU for an Energy Trader
Receive the existing simulator (C++17 + OpenMP, around 5,000 lines), the test book (around 8,000 spread options), and access to an H100 80GB. Port to CUDA: random-number generat…
- Gpu Programming
- Cuda
- Monte Carlo
Advanced Concurrency and Parallel Computing - ResearchSeniorNew
Benchmark Reward-from-Feedback Methods on a Tabletop Pick-Place
You will use a Franka Panda arm in PyBullet on a 4-object pick-and-place task. For each of the three feedback methods, train a reward model and a downstream policy until converg…
- Reinforcement Learning
- Reward Learning
- Preference Comparison
Human-Robot Interaction - ResearchSeniorNew
Benchmark Long-Context Architectures on a Legal-Doc Retrieval Task
You receive a public legal-QA dataset (e.g., LongBench's legal split or similar) filtered to documents over 50,000 tokens. Implement or wrap 3 architectures: a sliding-window Tr…
- Long Context Architectures
- State Space Models
- Transformers
Advanced Deep Learning 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
- CodeSeniorNew
Port a Microkernel's Fast Message Path from x86-64 to RISC-V
Starting from the representative fast-path module and its behavioral specification in the provided materials, re-implement the synchronous fast message path for RISC-V 64-bit (t…
- Microkernels
- Ipc
- Risc V
Open coursework - CodeSeniorNew
Assess Secure Cross-Bank Fraud Computation for a Four-Bank Consortium
Using the de-identified situation in the consortium case file and the two synthetic per-bank datasets (the bank suspicious-account lists and the merchant fraud-marker tallies), …
- Secure Computation
- Mpc
- Cryptography
Open coursework - DesignSeniorNew
Design an AVX-512 Vectorization Proposal for a Hot Simulation Kernel
Working only from the provided scalar kernel module, the benchmark dataset, the profiling readout, and the target-environment brief, propose and justify a vectorization design f…
- Vectorization
- Llvm
- Loop Optimization
Open coursework - DesignSeniorNew
Design a Lock-Free Concurrent Skip List for a Time-Series Database
Using the baseline implementation and benchmark harness in the provided code module (baseline-skiplist-rs), design and build a lock-free skip list that replaces the mutex-guarde…
- Lock Free
- Concurrent Data Structures
- Rust Programming
Open coursework - CodeSeniorNew
Build a Unikernel for a Privacy-Focused VPN Provider
Choose either MirageOS or Unikraft as the unikernel framework. Build an image running a minimal WireGuard data plane (existing OCaml or Rust WireGuard reference implementation i…
- Unikernels
- Virtualization
- Wireguard
Advanced Operating Systems - CodeSeniorNew
Profile and Cut Inference Cost on a Recommender at Scale
You receive (1) a frozen ONNX export of the production model, (2) a sample request trace of 24 hours at 1% sampling, and (3) a single A100-class GPU sandbox. Profile with NVIDIA…
- Gpu Profiling
- Model Quantization
- Inference Optimization
Open coursework - ResearchSeniorNew
Trajectory Prediction Model for Urban Robotaxis
Use the Argoverse 2 motion-forecasting dataset (open access). Train an LSTM baseline + a transformer challenger (e.g., a small Wayformer or HiVT). Evaluate on minADE/minFDE (min…
- Trajectory Prediction
- Transformer Models
- Evaluation
Open coursework - ResearchSeniorNew
Implement an Autoregressive Model for Anonymized Voice-Synthesis at a Defense Vendor
You receive a public-domain speech dataset (LibriTTS subset, around 50 speakers) and a fixed evaluation protocol (speaker-identifiability AUC, emotion-preservation MOS proxy, in…
- Autoregressive Models
- Voice Conversion
- Speech Synthesis
Deep Generative Models - CodeSeniorNew
Bulk Synchronous Parallel PageRank for a Web-Graph Lab
Working from the lab situation record (`lab-case-file`), implement 25 iterations of PageRank under a Bulk Synchronous Parallel (BSP) model using either Apache Spark with GraphX …
- Parallel Algorithms
- Bsp
- Graph Algorithms
Open coursework - CodeSeniorNew
Heterogeneous Scheduling Policy for a Big.LITTLE Mobile SoC
Read the existing scheduler's energy model (provided). Design an improved policy using one or more of: per-task IPC-aware placement, exponential moving average load smoothing to…
- Multicore Architecture
- Scheduling
- Energy Modeling
Advanced Computer Architecture - DesignSeniorNew
OTA Firmware-Update Architecture for a 50,000-Device Smart-Meter Fleet
Design a 4-stage OTA architecture: (1) signed image build + manifest, (2) backend rollout (1 percent canary, 10 percent expand, 50 percent expand, 100 percent), (3) device-side …
- Embedded Systems
- Cyber Physical Systems
- Edge Computing
Internet of Things and Cyber-Physical Systems - CodeSeniorNew
Design a Polyhedral Loop-Tiling Strategy for a Tensor DSL
Study the DSL's IR (provided, MLIR-based with linalg-style ops). Choose a tiling strategy (rectangular tiling with cost-modeled tile sizes is the safe baseline; full polyhedral …
- Polyhedral Analysis
- Loop Tiling
- Mlir
Advanced Compilers and Program Optimization
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.