Create project
Paste a real business goal, choose the workflow type, and let Hermes preserve the project brief as the source of truth.
Project examples
See what your AI operations team can actually do
Hermes is not limited to outreach. Start with a business objective, generate an agent topology, deploy it to a local runtime, and watch specialized tasks produce records, outputs, approvals, and delivery-ready work.
Five workflow examples. One operating model. Real pages, records, tasks, and outcomes.
Operating model
Every project follows the same productized path
Generate blueprint
Hermes creates supervisors, specialist workers, reviewers, prompts, schedules, artifact contracts, and quality gates.
Deploy locally
Publish the workflow from the cloud control center to a paired LocalConnector and Hermes runtime you control.
Run and inspect
Watch tasks move through Queued, In Progress, Completed, Failed, or Cancelled states while records and outputs are created.
Review and deliver
Use Outputs, Approvals, Records, Email Review, Deliveries, and Learnings to turn AI work into approved business outcomes.
Workflow catalog
Choose the example closest to your work
Outreach and GTM
Kickstarter backer discovery and creator outreachResearch and intelligence
Competitive intelligence brief for a new marketInternal operations
Weekly customer onboarding risk reviewContent and campaigns
Founder-led launch content engineGeneral AI operations
Executive AI operations command center
Outreach and GTM
Kickstarter backer discovery and creator outreach
Use Hermes as a campaign promotion team that researches high-fit backers, creators, communities, and partners, then prepares reviewable outreach instead of sending generic blasts.
Business outcome
A ranked pipeline of high-fit targets, public channels, personalized drafts, review notes, and daily campaign actions.
Create project with this description
Create a Kickstarter promotion project for Hermes Control Center.
Find high-fit early backers, AI automation creators, SaaS founders, agencies, newsletters, podcasts, and communities in the United States and Canada.
Score each target by relevance, public evidence, audience fit, safe contact channel, and likelihood to share or back the campaign.
Produce personalized email, LinkedIn, X, newsletter, podcast, and community post drafts.
Include the campaign link naturally, mention early access when appropriate, and keep every draft Pending Review until a human approves it.Where the work appears
Runs
Monitor target discovery, contact discovery, copywriting, review, and daily summary tasks.
Records
View qualified targets, source evidence, personalization signals, compliance notes, and rejected reasons.
Outreach Leads
Inspect ranked people, companies, creators, communities, and partners with fit scores.
Outreach Contacts
Review safe public contact channels and confidence notes.
Email Review
Approve, reject, or revise outbound drafts before anything is sent.
Outputs and Deliveries
Find community posts, creator pitches, promotion plans, and delivery-ready campaign assets.
Choose Outreach and GTM, paste the campaign brief, add target audiences, product highlights, proof points, geographic focus, and compliance rules.
Deploy to the active local node so discovery, scoring, and copywriting run through the paired LocalConnector.
Click Run now or use the daily schedule. The run creates discovery, channel, copywriting, review, and summary tasks.
Review drafts, approve the best ones, send only approved outreach, and use learnings to improve the next run.
Run task plan
What the run creates
Coordinate daily promotion mission
Break the campaign goal into discovery, channel, copy, review, and reporting work.
Output: Daily mission plan and top priorities.Find and score backers/promoters
Search public sources for relevant founders, creators, agencies, newsletters, podcasts, and communities.
Output: Qualified targets with fit scores and source URLs.Find safe contact routes
Identify public emails, forms, profiles, and community participation paths without guessing private data.
Output: Contact channels with confidence and risk notes.Write conversion-oriented drafts
Use target evidence plus product positioning to produce specific, respectful outreach.
Output: Pending Review email, LinkedIn, X, creator, and community drafts.Protect brand and compliance
Reject generic, spammy, unsupported, or risky drafts before approval.
Output: Review decisions, revision notes, and approved assets.
How the operator reaches the final goal
Open Email Review for outbound drafts, Outputs for creator and community assets, Outreach Leads for ranked targets, Outreach Contacts for safe channels, and Learnings for next-day targeting improvements.
Research and intelligence
Competitive intelligence brief for a new market
Use Hermes as a research analyst team that turns broad questions into sourced findings, ranked competitors, open questions, and an executive-ready decision memo.
Business outcome
A sourced market map, competitor profiles, confidence-scored findings, and a decision memo your team can act on.
Create project with this description
Create a research intelligence project for a B2B SaaS company evaluating expansion into Canadian SMB accounting automation.
Research competitors, buyer pains, pricing patterns, integration expectations, content gaps, community discussions, and partnership opportunities.
Capture source URLs, confidence levels, risks, and open questions.
Produce a concise executive brief, ranked recommendations, and a next-research plan.
Separate evidence from inference and mark unsupported claims clearly.Where the work appears
Runs
Track source discovery, competitor profiling, synthesis, and evidence review tasks.
Records
Review findings, entity profiles, source notes, open questions, risks, and confidence levels.
Outputs
Open research briefs, decision memos, ranked recommendations, and next-step plans.
Approvals
Approve decision memos or request changes before leadership uses them.
Learnings
Capture better search queries, source patterns, and rejected assumptions for the next research pass.
Choose Research and intelligence, define the decision question, target market, source quality rules, and what a useful final memo must answer.
Deploy the blueprint to a node with access to the local runtime and any allowed research tools or source files.
Run manually for a one-time study or schedule recurring monitoring. Tasks collect sources, extract findings, synthesize patterns, and review evidence.
Use the final brief to choose market priorities, assign follow-up research, and approve recommendations.
Run task plan
What the run creates
Frame the decision question
Convert the brief into research questions, source priorities, and acceptance criteria.
Output: Research plan and task assignments.Collect evidence
Find public sources, summarize credibility, and capture limitations.
Output: Source notes and evidence records.Profile market players
Compare positioning, features, pricing signals, audiences, and go-to-market patterns.
Output: Entity profiles and competitor comparison.Turn evidence into insight
Cluster findings, identify implications, score opportunities, and name unknowns.
Output: Ranked recommendations and open questions.Check quality
Validate source trail, confidence, unsupported claims, and missing context.
Output: Reviewed decision memo ready for approval.
How the operator reaches the final goal
Open Records to audit evidence, Outputs for the executive brief, Approvals to accept the memo, and Learnings to schedule the next intelligence cycle.
Internal operations
Weekly customer onboarding risk review
Use Hermes as an operations coordinator that reviews active work, identifies blockers, prepares action items, and packages sensitive decisions for approval.
Business outcome
A weekly operating packet with blockers, owners, risks, action items, handoff notes, and approval-ready decisions.
Create project with this description
Create a weekly customer onboarding operations project.
Review the onboarding pipeline, identify accounts at risk, summarize blockers, propose owners, draft action items, and prepare escalation notes.
Generate a concise status summary for the leadership team and a handoff note for customer success.
Mark any customer-facing or escalation-sensitive action as Pending Review.
Keep outputs practical, owner-oriented, and usable in the next team meeting.Where the work appears
Runs
Watch risk triage, blocker review, action planning, runbook drafting, and approval review tasks.
Records
Review work items, blockers, decisions, risks, owners, priority, and due dates.
Outputs
Open status summaries, handoff notes, runbooks, and approval packets.
Approvals
Review escalation notes, customer-facing updates, and sensitive operational recommendations.
Learnings
Capture recurring blockers and process improvements for future runs.
Choose Internal operations, define the operating cadence, owner rules, risk criteria, and what should require approval.
Deploy after connecting the local runtime to the operating context or source documents your team allows it to inspect.
Run before the weekly meeting. The system creates risk review, blocker triage, action planning, runbook, and approval packet tasks.
Use the packet in the meeting, approve escalation notes, and turn action items into follow-up work.
Run task plan
What the run creates
Own the operating cadence
Assign workstream reviews, consolidate status, and enforce approval-sensitive rules.
Output: Weekly operating plan and summary.Inspect the work queue
Identify stalled items, missing inputs, owner gaps, and high-risk accounts.
Output: Work items, blockers, and risks.Create next steps
Turn blockers into owner-specific actions with priority and acceptance criteria.
Output: Action items and handoff notes.Make the process repeatable
Document checklists, escalation paths, and handoff patterns.
Output: Runbook and SOP updates.Review sensitive decisions
Check escalation language, customer-facing risks, and decision readiness.
Output: Approval packet and review notes.
How the operator reaches the final goal
Open Outputs for the status packet, Records for blockers and owners, Approvals for sensitive actions, and Learnings to improve the next weekly run.
Content and campaigns
Founder-led launch content engine
Use Hermes as a content operations team that researches audience pains, creates angles, drafts channel-specific assets, and keeps publishable content in review.
Business outcome
A campaign package with audience insights, content briefs, post drafts, channel variants, editorial risks, and a reviewable calendar.
Create project with this description
Create a founder-led launch content project for a new AI operations platform.
Research the audience of SaaS founders, operators, agencies, AI builders, and productivity tool users.
Create content angles around workflow automation, local execution, approvals, research, outreach, and real business outcomes.
Draft LinkedIn posts, X posts, community posts, newsletter copy, founder story posts, and demo CTAs.
Keep every publishable asset Pending Review and flag unsupported claims or tone risks.Where the work appears
Runs
Track audience research, angle creation, asset drafting, editing, and final campaign packaging tasks.
Records
Review audience insights, content angles, channel findings, and editorial risks.
Outputs
Open content briefs, post drafts, campaign packages, and editorial calendars.
Approvals
Approve or request changes before publishing assets.
Deliveries
Route approved social posts, newsletter snippets, and campaign assets to the next publishing step.
Learnings
Capture which angles, channels, and CTAs should be tested next.
Choose Content and campaigns, provide audience segments, product positioning, proof points, forbidden claims, channels, and desired publishing cadence.
Deploy to the local node so research, drafting, channel adaptation, and editorial review run as one workflow.
Run daily during the launch window. Tasks generate audience insights, content angles, drafts, variants, and editorial review notes.
Approve publishable drafts, route final assets, and use learnings to improve hooks and channels.
Run task plan
What the run creates
Own the campaign mission
Break the launch into audience, angle, asset, review, and delivery work.
Output: Daily campaign plan.Find pains and motivations
Identify objections, desired outcomes, channel context, and buyer language.
Output: Audience insights and channel findings.Create angles
Turn audience evidence and product strengths into campaign narratives.
Output: Content angles and brief outlines.Draft channel-specific copy
Write posts, newsletters, demo CTAs, community posts, and founder notes.
Output: Reviewable post drafts and campaign package.Raise quality
Check clarity, claims, voice, specificity, and channel fit.
Output: Approved drafts or change requests.
How the operator reaches the final goal
Open Outputs for briefs and drafts, Approvals for publish-ready content, Deliveries for final assets, and Learnings for the next content iteration.
General AI operations
Executive AI operations command center
Use Hermes for mixed workflows that combine research, planning, decision support, content, approvals, and delivery without forcing everything into an outreach template.
Business outcome
A custom operating system for the business process you describe: records, work products, recommendations, tasks, reviews, and learning loops.
Create project with this description
Create an AI operations command project for the executive team.
Each run should inspect the current strategic objective, break it into specialist tasks, gather findings, create decision inputs, propose action items, produce a concise operator brief, and flag anything that needs human approval.
Do not assume this is an outreach project unless the brief explicitly asks for outreach.
Match records and outputs to the objective, keep recommendations grounded, and create follow-up tasks when more work is required.Where the work appears
Runs
Inspect supervisor coordination, specialist tasks, review tasks, and follow-up work.
Records
View findings, action items, decision inputs, risks, opportunities, and objective-specific records.
Outputs
Open briefs, work products, recommendations, tables, checklists, memos, or plans.
Approvals
Review sensitive recommendations and decision inputs before execution.
Learnings
Capture patterns that should change future topology, prompts, or quality gates.
Choose General AI operations for custom or mixed workflows, then describe the operating goal, acceptance criteria, output format, and review rules.
Deploy once the topology looks right. The supervisor and specialist agents will follow the brief rather than a fixed GTM schema.
Run on demand or on schedule. The run creates coordination, specialist, review, synthesis, and follow-up tasks based on the project objective.
Use Records, Outputs, Approvals, Runs, and Learnings as the command layer for ongoing AI-assisted business execution.
Run task plan
What the run creates
Translate the brief into work
Read the objective, create the operating plan, and assign specialist tasks.
Output: Mission plan and task graph.Do domain-specific work
Research, analyze, draft, compare, summarize, or operationalize the exact brief.
Output: Findings, work products, and action items.Connect the pieces
Combine task outputs into a concise business deliverable.
Output: Brief, recommendation, or decision input.Check usefulness
Review evidence, completeness, risks, and whether the output is ready for use.
Output: Review notes and approval packet.Improve the next run
Identify missing context, better tasks, recurring issues, and prompt improvements.
Output: Learning notes and follow-up recommendations.
How the operator reaches the final goal
Open Runs to inspect execution, Records for reusable operational memory, Outputs for business deliverables, Approvals for human decisions, and Learnings to tune the workflow over time.