Full Workflow

By Felipe Thomaz Pedroni

Precedes is a optimization-first platform where software optimization is delivered as a structured service — with defined goals, measurable results, and payment tied directly to verified improvement. No ratings. No reviews. Just math.

🔁 Step-by-Step Workflow

🧑‍💼 For Clients

Explore Optimization Services

Clients browse a curated catalog of optimization services offered by vetted engineers. Each service specifies:

• Architecture target: CPU, GPU, TPU, FPGA
• Supported languages: C, C++, Python, Fortran
• Optimization metrics: Runtime, memory usage, throughput
• Optimization tiers: Light, Moderate, Aggressive — with transparent pricing

Select an Optimizer

Clients choose an optimizer based on expertise and architecture focus. Every optimizer displays a public score, based entirely on past optimization — not opinions.

Upload Code or Grant Access

Clients securely share their environment or codebase using:

• Private Git repositories
• Containerized environments
• Preconfigured sandbox instances

Agree on Tier and Metric

Client and optimizer finalize:

• Tier: Light (3–5 tests), Moderate (10), Aggressive (20+)
• Metric: Execution time, memory footprint, throughput, etc.
• Baseline: Captured and frozen before any modification

Optimization Begins

The optimizer works within the agreed scope and deadline. All iterations are transparently logged and visible to the client.

Review and Verification

Precedes verifies the delivered optimization using the exact same environment. Tests confirm:

• Percentage of promised improvement achieved
• Statistical consistency (e.g. average, standard deviation)
• Results are locked. Disputes are avoided. The math speaks for itself.

Payment by Optimization

Payment is proportional to the actual gain. If 80% of the goal is achieved, only 80% of the price is charged. Precedes applies a 10% platform fee — only to the optimizer.

👨‍💻 For Freelancers (Optimizers)

Create Your Optimization Service

• Optimizers publish a service specifying:
• Target architectures and supported languages
• Optimization metric types
• Tier-based pricing model

Receive Orders

Once selected, the optimizer reviews the codebase, captures the initial benchmark, and defines the optimization strategy.

Work with Transparency

All progress is tracked in-platform. Iterations are visible. Communication with the client is streamlined — no off-platform negotiation.

Submit Final Results

Upon delivery, Precedes reruns the benchmarks under the same test conditions and compares against the frozen baseline.

Get Paid by Objective Results

The optimizer is paid in direct proportion to the gain achieved. Their score is updated using:

🛡️ Security and Fairness

• All freelancers are KYC-verified
• Code is never published or shared
• No subjective reviews — only results matter
• Every service includes: metric, deadline, and success condition
• This isn’t freelancing. This is optimization with precision.