# Live Builds

Canonical page: https://antern.co/live-builds/

## Summary

Antern participants build systems, not disposable projects.

Every build exists to teach a discipline: context, agents, retrieval, evaluation, reliability, product judgment, or outreach systems.

The public page describes build families and engineering standards. Exact project briefs, datasets, and weekly specs are shared through the counsellor and cohort flow.

## Build Families

### AI Coding Systems

Participants study coding agents as systems: context loading, planning, edits, tests, review, rollback, logs, and human approval before risky changes.

### Retrieval + Knowledge Systems

Participants build ingestion, chunking, embeddings, hybrid search, reranking, context assembly, citation behavior, and retrieval evaluation.

### Agent Reliability Harnesses

Participants design agents that can handle ambiguity, missing capabilities, policy boundaries, tool failures, token budgets, and inconsistent user intent.

### Evaluation Platforms

Participants create datasets, slices, judges, regression gates, traces, dashboards, and review workflows for AI systems that must improve safely.

### AI Product Systems

Participants connect model behavior to product workflows: user intent, backend services, queues, observability, cost, latency, and deployment.

### Outreach + Operator Systems

Participants learn to build AI-assisted systems for research, ICP design, campaign debugging, personalization, CRM hygiene, and opportunity creation.

## Build Lifecycle

1. Frame: define the user, task, constraints, risks, and what success would actually mean.
2. Design: write architecture, invariants, interfaces, context strategy, and evaluation plan before coding.
3. Build: implement with AI assistance while preserving checkpoints, logs, and human decision ownership.
4. Evaluate: run tests, slices, traces, critics, regressions, and failure probes against the system.
5. Defend: explain tradeoffs, rejected alternatives, limits, cost, latency, and what would fail in production.
6. Publish: turn the build into proof-of-work: demo, writeup, architecture note, and evaluation report.

## Hard Constraints

Participants do not only build happy-path demos. They learn how systems behave when tools are missing, context is incomplete, policies constrain behavior, costs matter, latency matters, and a human must approve risky actions.

Hard constraints include:

- Ambiguous or incomplete user requests
- Missing tools or unsupported capabilities
- Policy-constrained actions
- Token and latency budgets
- Expensive model calls
- Context overflow
- Tool failures and retries
- Human approval gates

## Capstone Directions

Final systems are chosen for depth, defensibility, and signal.

Examples of public capstone directions include PR review agents, research agents, knowledge agents, sales or outreach agents, agent observability platforms, and evaluation harnesses.

Every serious capstone must include architecture, evaluation, deployment thinking, and proof-of-work.

## Project Details

Exact weekly projects, datasets, deliverables, evaluation rubrics, and expected proof-of-work are shared through the Antern counsellor and cohort flow.