NervScale AI helps applied AI teams plan, evaluate, and ship with confidence

A focused workspace for the work behind reliable AI systems.
Keep evaluations, preference data, and release plans moving forward.

NervScale AI
Search...⌘K
Inbox3
My Issues
Workspace
Initiatives
Projects
Views
Teams
Favorites
Mobile App
Q1 Roadmap
API Docs
Your Teams
Product
Engineering
Help & Support

Inbox

EVAL-135

Review Terminal-Bench run

nan assigned you

2h
EVAL-42Project

SWE-bench Pro regression

New project update by raissa

1d
BENCH-159

SpreadsheetBench import delay

Run is delayed

2d
OPS-498

Update reviewer guidelines...

karri mentioned you

4h
REL-160

Release gate is blocked

You asked to be reminded

1w
Project

AppWorld score review

Added as project member

DOC-122

Benchmark docs update

erin marked as Duplicate

1w
Project

Release readiness

New project update by paco

QualityTerminal-BenchEVAL-135

Review Terminal-Bench run

Run.reviewContext is missing from this run. It should include a BenchmarkResult record.
@BenchmarkRun(BenchmarkResult,reviewers,status:ready)
public reviewContext?: BenchmarkResult;
We would compareBenchmarkResult.reviewContextbefore the release is cleared.
Add follow-ups
Links
#20319 kal/eval-135 add reviewer context to benchmark run
Activity

nan moved from Backlog to In Progress

5 months ago

alex commented on this issue

5 months ago

Benchmark coverage

The benchmarks behind reliable release decisions.

Code, terminal, and spreadsheet workflows in one operating view.

Terminal-Bench
SWE-bench Pro
SWE-bench
SpreadsheetBench
Terminal-Bench
SWE-bench Pro
SWE-bench
SpreadsheetBench
Terminal-Bench
SWE-bench Pro
SWE-bench
SpreadsheetBench
Terminal-Bench
SWE-bench Pro
SWE-bench
SpreadsheetBench

Made for modern product teams

NervScale AI is shaped by the practices and principles that distinguish world-class product teams from the rest: relentless focus, fast execution, and a commitment to the quality of craft. Make the switch

Applied AI operations

One workspace for evaluation operations

NervScale AI for teams. Keep evaluation, preference-data, and release work connected without losing ownership or context.

Learn more
Route work to...
EvaluationTeam
Data qualityTeam
ResearchTeam
PlatformTeam
SafetyTeam
OperationsTeam

Cross-functional evaluation operations

Keep RLHF queues, evaluation runs, and release context clear from planning through launch.

Team routing
Suggestions
nanMobile App RefactorSlack
Duplicate of
Related to
nan

Why this owner was selected

This person owned prior preference-data reviews for the same evaluation workflow and release window.

Alternatives

yannerin
Accept suggestion

Evaluation visibility

Keep evaluation runs, pull requests, and launch decisions connected to the work they affect.

//evals.nervscale.ai/events

"workflows": {

"evaluationRuns": {

"command": "npx"

Add an update
Evaluation and release planning

Set the product direction

Align your team around a unified evaluation timeline. Plan, manage, and track preference data, model quality, and releases with NervScale AI's visual planning tools.

30
AUG 3
10
17
AUG 22
24
SEP
Release coordination
Prototype
Beta
RLHF preference review

Manage projects end-to-end

Consolidate specs, milestones, tasks, and other documentation in one centralized location.

Project Overview

Properties
In ProgressENG
Resources
🚩ExplorationUser interviews
Milestones
Design Review 100%Internal Alpha 100% of 10GA 25% of 53

Project updates

Communicate progress and project health with built-in project updates.

Off track
At risk
On track

We are ready to launch next Thursday

Sep 8

Ideate and specify
what to build next

Collaborative documents
Inline comments
Text-to-issue commands
Spice harvesterProject specs•••
Collaborate onideaszoe

Write down product ideas and work together on feaquinnture specs in realtime, multiplayer project documents. Add **style** and ##structure with rich-text formatting options.

Initiatives

Coordinate strategic product efforts.

Cross-team projects

Collaborate across teams and departments.

Milestones

Break projects down into concrete phases.

Progress insights

Track scope, velocity, and progress over time.

Workflows and integrations

Collaborate across
tools and teams

Expand the capabilities of the NervScale AI system with a wide variety of integrations that keep everyone in your organization aligned and focused.

Intercom·zoe@acme.inc

We need a cost breakdown across...

A
ACMENew request
Multi-cloud costdashboard
Planning
📅Q4 2025

Customer Requests

Build what customers actually want

#20319igor/lin 15287add sourc...
igorlinkedigor/lin 15287add sou...
igorchanged status from In Progre...
GitHubchanged status from In Revie...
igorchanged status from Ready...

Powerful git workflows

Automate pull requests and commit workflows

Inbox

Mobile workspace

Move product work forward from anywhere

Request intake

Turn workplace requests into actionable issues

Integrations

Keep every engineering tool connected

Figma Integration

Bridge the gap between engineering and design

SPRINT API

Built for developers

Build your own add-ons with the NervScale API

Plan the present. Build the future.