Competitive Positioning
The AI workspace that orchestrates.
Every other tool helps one person do one thing. Obvious runs the whole team.
Where Obvious sits
Map any AI product on two axes.
Breadth — how many workflows it covers. Depth — how completely it executes. Three quadrants are commoditizing, consolidating, or collapsing. One holds.
Low breadth · High depth
Point Solutions
Deep in one lane. Legal, code, revenue, design, and nothing outside it. Complementary to Obvious; the typical team runs both.
Harvey · Cursor · Gong · Claude Code · Figma AI
High breadth · High depth
Full Workflow AIObvious
Executes across every workflow. Produces connected artifacts. Runs on schedule. Compounds knowledge over time.
Obvious · Manus · Claude Agents · OpenAI
Low breadth · Low depth
LLM Wrappers
One task on top of someone else's model.
Jasper · Copy.ai · Writesonic · Otter.ai
High breadth · Low depth
Enterprise LLMs
Answers in a chat window. The user still does the work.
ChatGPT · Gemini · Copilot · Glean · Perplexity
What Obvious delivers
Eight capabilities that separate a real workspace from a chat window.
Some are table stakes you’d expect. A few aren’t.
Multiplayer by default
A shared workspace with four permission levels, inline comments, real-time collaboration, and a project README that keeps the agent briefed on how your team works.
Server-side VM execution
Every job runs on our infrastructure — fast on large workloads, consistent on repeated ones. Auto-checkpoints snapshot the project before any destructive operation. Roll back anything, in plain language.
Enterprise architecture
SOC 2 compliant on infrastructure that predates most AI companies. SSO via SAML, exportable audit logs, IP access rules, content filtering policies, and per-role AI mode restrictions.
Model selection, handled
Obvious evaluates foundation models for what they're actually good at — so you don't run your own evals every time one ships. A dozen+ modes route each task to the right model. Custom Modes let you add your own grounding.
14+ native artifact types
Docs, Workbooks, six live-connected views (Kanban, Calendar, Timeline, Gallery, Checklist, Form), Folios, Dashboards, Apps, Recordings, Tasks, Canvas, and Images. The agent produces the thing, not a summary of it.
Custom Modes & Skills
Set your own system prompt. Connect any API, MCP, or internal service via secrets management. Skills encode your team's workflows. The intelligence is yours to define.
Connect to anything
Native connectors for Slack, Gmail, Outlook, Google Calendar, Salesforce, HubSpot, Attio, Notion, Granola, and GitHub are built in. Anything else connects via secrets management. If it has an endpoint, Obvious can reach it.
Automations
Scheduled tasks run daily, weekly, or monthly. Webhooks trigger agent workflows from any external event — a new Salesforce record, a Stripe payment, a Slack message. Approval gates pause execution for human review.
A workspace that compounds
Most AI tools give you a better start. Obvious gives you a better run.
Every task the agent handles makes the next one faster and more specific to how your team works.
Memory at every layer
User memory tracks preferences and working style. Project memory holds decisions, schemas, and standing context. Workspace memory extends across the whole organization. The agent knows more each session — no one briefs it from scratch. No one writes the same summary twice.
Skills encode your best work
When your team figures out how to run a great analysis, a clean migration, or an exec-ready report, encode it as a Skill. Every agent in the workspace runs it the same way from then on. Institutional knowledge that doesn't walk out the door.
Live artifacts, not static exports
Workbooks connect to dashboards. Dashboards connect to reports. When the underlying data changes, everything downstream updates automatically. The Q2 pipeline summary isn't a file that goes stale — it's a living artifact the whole team reads from.
Automation that runs without you
Scheduled tasks execute on whatever cadence you set. Webhooks trigger workflows the moment something changes in Salesforce, Slack, or your own systems. The pipeline that used to live in someone's head runs on its own.
vs. the field
Honest comparisons based on live testing.
We ran the same 8 tasks across every tool and reported what happened.
vs
Manus
Manus is a capable autonomous agent. Obvious is a workspace. The difference is whether your work compounds or resets.
See comparison
vs
Notion
Notion is where you document what happened. Obvious is where the work actually happens.
See comparison
vs
Claude CoWork
Claude CoWork automates what's in your files. Obvious builds on what's in your workspace.
See comparison
vs
Google Gemini
Google Workspace is optimized for sharing and co-editing. Obvious is optimized for doing.
See comparison
vs
Claude Sonnet 4.6
Claude Sonnet 4.6 is a conversation. Every session starts from zero. Obvious is a workspace — the agent knows what you've already figured out.
See comparison
vs
ChatGPT
ChatGPT answers questions. Obvious does work. The distinction matters when the work needs to persist.
See comparison
The questions worth asking