Field guide 001 · rebuilt for practiceDecide → Build → Verify → Operate

Learn AI by shippingwork that survives review.

Work from a constrained field situation and finish with an artifact another person can inspect: a source ledger, workflow, eval set, approval policy, trace, or release gate.

Start with AI in MarketingCompare all tracks
Two collaborators reviewing a workflow map, source cards, and an approval checklist on a dark workbench.
Editorial illustration. Technical states and results appear inside the lessons as reproducible artifacts.
TRACKS
03 ordered paths
LESSONS
33 case-led builds
LOCALES
English + 繁體中文
PROOF
Artifacts and checks

Choose a track

Start with the system you need to run.

The marketing path starts from a business workflow. Agent Building starts from system boundaries. Harness Engineering starts from the repository around a coding agent. All three use named scenarios, explicit limits, failure clinics, and production checks.

Track 0110 MODULE

Operating track

AI Marketing

Build one marketing workflow that can be inspected, evaluated, recovered, and improved in production.

OutputYou will leave with a versioned operating packet: workflow charter, evidence ledger, content state model, search experiment, creative and lifecycle controls, replay-safe automation, agent mandate, evaluation scorecard, and 90-day deployment plan.Effort16 to 20 hours plus one bounded workplace pilot
Open the field guide
Track 0212 MODULE

For engineers, technical operators, and AI product owners

Agent Building

Build an agent you can test, interrupt, inspect, and operate.

OutputA bounded marketing research agent with claim-level evidence, explicit permissions, resumable human review, a regression suite, an operating cost model, and a documented go or no-go decision.Effort18 to 24 hours plus the capstone
Open the field guide
Track 0311 MODULE

For software engineers, technical leads, and developer-platform owners

Harness Engineering

Turn a coding agent into a repository system that can understand, verify, resume, and hand off real work.

OutputA measured repo harness with instructions, state, isolation, mechanical boundaries, end-to-end verification, run receipts, handoff, and a documented go or no-go decision.Effort25 to 30 hours including the 5 to 7 hour capstone
Open the field guide

Roadmap preview

Three paths, one operating standard.

Open any lesson directly, or follow the sequence so each artifact becomes an input to the next build.

AI Marketing

TRACK 01
  1. 00AI-Native Marketing FoundationsFoundation · 70 min
  2. 01AI Research and Customer IntelligenceLab · 85 min
  3. 02AI Content OperationsLab · 90 min
  4. 03SEO, GEO, and AI SearchLab · 95 min
  5. 04Creative and Performance MarketingLab · 85 min
View all modules · 10

Agent Building

TRACK 02
  1. 00AI agents vs workflows: earn the loopFoundation · 75 min
  2. 01LLM APIs and structured outputs: build a typed boundaryLab · 105 min
  3. 02Tool use and function calling: keep execution outside the modelLab · 120 min
  4. 03Context, memory, and state: make the job resumableLab · 125 min
  5. 04RAG and knowledge retrieval: test the evidence layer firstLab · 150 min
View all modules · 12

Harness Engineering

TRACK 03
  1. 00Diagnose the harness gap: measure before you scaffoldFoundation · 75 min
  2. 01Write an executable task contract: give done a truth valueLab · 100 min
  3. 02Make the repository the system of record: give agents a mapLab · 110 min
  4. 03Bootstrap and isolate the environment: make every run start from truthLab · 120 min
  5. 04Persist state and hand off cleanly: survive a fresh contextLab · 115 min
View all modules · 11

Choose by responsibility

Enter through the decision you already own.

01

Marketer

Turn one recurring marketing process into a measurable, reviewable workflow.

Start the marketing path02

Builder

Build tool-using systems that handle bad inputs, denied actions, and recovery.

Start the agent path03

AI lead

Set evaluation, permissions, rollout, rollback, and operating ownership before scale.

Review production readiness04

Harness engineer

Make a coding agent understand, verify, resume, and hand off repository work safely.

Build the repository harness

Featured builds

Leave with operating artifacts.

Evidence system01

Build a cited market brief

Normalize a fixed source pack, expose conflicting evidence, and verify key claims before synthesis.

Open the lab →
Content system02

Run a content state machine

Move one article through evidence, review, publication, and refresh with observable transition gates.

Open the lab →
Capstone03

Evaluate a research agent

Combine tools, state, retrieval, approval, traces, and a cost ceiling against a fixed benchmark.

Open the capstone →
Repository system04

Retrofit a brownfield codebase

Build the contracts, state, isolation, rules, evals, recovery, and evidence for a coding-agent task.

Open the harness capstone →

Tenten tools and field guides

Use the right asset at the next decision.

The Academy connects existing calculators, model comparisons, playbooks, and delivery methods to a concrete step in the work.

Service

Forward Deployed Marketing

Tenten's field approach to building and operating bounded AI marketing workflows.

Open resourceTool

AI model leaderboard

Compare current models, then validate the shortlist on your own task suite.

Open resourceField guide

Harness Engineering

The runtime, control, observability, and recovery layer around an agent loop.

Open resourcePlaybook

FDM Enterprise Playbook

A decision guide for scope, governance, operating ownership, and transfer.

Open resource

When the lab meets reality

Bring one artifact and one unresolved production decision.

We can review the evidence, permissions, evaluator, handoff, and rollback plan before you commit to a larger implementation.

Book an artifact reviewSee how Tenten delivers