# Confident Prep Full AI Search Context

Free, mentor-curated AI engineering and AI interview-prep learning paths with real, deployable projects.

## Positioning

Confident Prep is for engineers, students, and professionals who want interview-ready AI/ML engineering skill through deployable projects rather than passive courses. The site is free, mentor-curated, and linked to an open-source project library.

## Conversion Paths

- [Start with the live cohort](https://confidentprep.com/live/): urgent free AI/ML interview-prep cohort.
- [Pick a learning path](https://confidentprep.com/paths/): choose a project sequence by current background.
- [Read the AI interview hub](https://confidentprep.com/interview/ai/): scenario-based AI interview questions and model answers.
- [Subscribe from any blog post](https://confidentprep.com/blog/): fortnightly project delivery by email.

## AI Engineering Fundamentals

You're newer to engineering and want to get into AI/ML. You don't need ML math — you need to build and deploy real things.

- Audience: 12 projects · Starts from Python
- Outcome: Outcome: junior AI engineer–ready
- Path URL: https://confidentprep.com/paths/ai-engineering-fundamentals/

### Projects

- [Document Chat](https://confidentprep.com/paths/ai-engineering-fundamentals/document-chat/): A CLI tool that loads a PDF or text file and answers questions about it, with streaming output and exact source citations for every answer.
- [Structured Output Extractor](https://confidentprep.com/paths/ai-engineering-fundamentals/structured-output-extractor/): Extracts structured data from unstructured text — job postings, meeting notes, support tickets — into validated Pydantic objects, with automatic retry when the LLM returns malformed JSON.
- [RAG Pipeline](https://confidentprep.com/paths/ai-engineering-fundamentals/rag-pipeline/): Embeds a folder of documents into a persistent FAISS index and answers questions by retrieving semantically relevant chunks, with score-threshold filtering and an A/B mode that puts RAG and no-RAG answers side by side.
- [Context Window Manager](https://confidentprep.com/paths/ai-engineering-fundamentals/context-window-manager/): A multi-turn CLI chatbot that handles context window limits with two strategies — sliding window and summarisation compression — and shows you exactly what each approach loses.
- [Multi-step LLM Pipeline](https://confidentprep.com/paths/ai-engineering-fundamentals/multi-step-llm-pipeline/): A 3-stage pipeline that extracts entities from any topic, enriches them with live Wikipedia data, and synthesises a professional briefing document.
- [Tool-Calling Agent](https://confidentprep.com/paths/ai-engineering-fundamentals/tool-calling-agent/): An agent that uses tools — calculator, datetime, and Wikipedia — to answer questions it can't answer from training alone, with a reasoning trace and max-iterations guard.
- [Persistent Agent with Memory](https://confidentprep.com/paths/ai-engineering-fundamentals/persistent-agent/): A CLI chatbot that remembers you. It stores conversation history and key facts in SQLite and recalls them across sessions — your name, preferences, and past context are available every time you return.
- [Streaming Chat API](https://confidentprep.com/paths/ai-engineering-fundamentals/streaming-chat-api/): A FastAPI server that streams LLM responses over SSE, tracks token usage and cost per request in SQLite, and rate-limits callers per IP.
- [Prompt Evaluation Framework](https://confidentprep.com/paths/ai-engineering-fundamentals/prompt-eval/): A test harness for LLM prompts. Define test cases in YAML, score responses with an LLM judge across multiple dimensions, and get exit code 1 on failure — so you can run it in CI.
- [Dockerize Your App](https://confidentprep.com/paths/ai-engineering-fundamentals/dockerize-llm-app/): Containerize a streaming chat API with a multi-stage Dockerfile, a persistent SQLite volume, and a health check.
- [Deploy to AWS Lambda](https://confidentprep.com/paths/ai-engineering-fundamentals/aws-lambda-deploy/): Deploy a containerised FastAPI chat API to AWS Lambda with least-privilege IAM, Secrets Manager, and a CloudWatch error-rate alarm.
- [Full AI Assistant (Capstone)](https://confidentprep.com/paths/ai-engineering-fundamentals/ai-assistant-capstone/): A deployed AI assistant combining RAG over uploaded documents, tool use for live data, context window management, and streaming responses — with an eval suite that runs against the live endpoint.

## AI-Augmented Engineering

You have 2–8 years of software experience. AI is everywhere and you feel behind. These projects put AI into the work you already know how to do.

- Audience: 12 projects · Any language welcome
- Outcome: Outcome: AI in your toolkit
- Path URL: https://confidentprep.com/paths/ai-augmented-engineering/

### Projects

- [Code Review Bot](https://confidentprep.com/paths/ai-augmented-engineering/code-review-bot/): A CLI that reads a git diff, chunks it per file, and uses an LLM in JSON mode to return prioritized, structured findings by severity and category.
- [PR Diff Summarizer](https://confidentprep.com/paths/ai-augmented-engineering/pr-diff-summarizer/): Turns a git diff into a plain-English summary, an architecture impact paragraph, and a test coverage flag, with zero raw code lines in the output.
- [Semantic Codebase Search](https://confidentprep.com/paths/ai-augmented-engineering/semantic-codebase-search/): Indexes a Python codebase at function granularity using AST chunking and FAISS, skips unchanged files via file hashing on re-index, and compares two embedding models side by side on the same query.
- [Code Documentation Generator](https://confidentprep.com/paths/ai-augmented-engineering/code-doc-generator/): Parses Python files with the ast module to extract function signatures and type hints, then generates LLM-written README documentation for every public function, with a diff mode that previews exactly what changed.
- [LLM Test Generator](https://confidentprep.com/paths/ai-augmented-engineering/llm-test-generator/): Generates pytest test cases for Python functions using an LLM, then validates them by actually running pytest — retrying with error context on failure.
- [LLM Observability](https://confidentprep.com/paths/ai-augmented-engineering/llm-observability/): OpenTelemetry-based observability wrapper for LLM calls — records model, tokens, latency, and cost per call, flags anomalies, exports to terminal dashboard or Jaeger.
- [Incident Summariser](https://confidentprep.com/paths/ai-augmented-engineering/incident-summariser/): Converts noisy log files (up to 100MB) into structured incident reports with timeline, root cause, and action items — using pre-filtering to reduce token usage before any LLM call.
- [Team Knowledge Extractor](https://confidentprep.com/paths/ai-augmented-engineering/team-knowledge-extractor/): Extracts structured decisions from GitHub issues and ADR markdown using an LLM, then ranks search results with keyword + recency scoring instead of embeddings.
- [Docs Copilot](https://confidentprep.com/paths/ai-augmented-engineering/docs-copilot/): Metadata-aware RAG documentation assistant with section-level citations, freshness tracking, and an "I don't know" confidence threshold.
- [SQL Agent](https://confidentprep.com/paths/ai-augmented-engineering/sql-agent/): Natural language to SQL agent with sqlglot AST-based safety validation (SELECT-only), schema introspection, and a self-correcting retry loop that injects SQL errors back into the prompt.
- [AI CLI Toolkit](https://confidentprep.com/paths/ai-augmented-engineering/ai-cli-toolkit/): Plugin-based ai command that unifies code review, PR summarizing, semantic search, and NL-to-SQL into one dynamically-loaded toolkit, gated by a YAML config.
- [AI PR Reviewer (Capstone)](https://confidentprep.com/paths/ai-augmented-engineering/ai-pr-reviewer/): GitHub Action that runs code review and test generation on every PR, posts findings as comments, enforces a severity gate, and tracks cost/latency via OpenTelemetry.

## ML Engineering on AWS

You've done some ML or data science and want to own the full system — training, serving, monitoring — in production on AWS.

- Audience: 12 projects · Python + ML exposure
- Outcome: Outcome: own ML systems end-to-end
- Path URL: https://confidentprep.com/paths/ml-engineering-on-aws/

### Projects

- [Local to SageMaker](https://confidentprep.com/paths/ml-engineering-on-aws/local-to-sagemaker/): Trains an XGBoost classifier on the UCI Adult dataset locally with a CLI script, then submits the identical script as a managed SageMaker training job via a least-privilege CDK-provisioned IAM role, with a cost estimator and CloudWatch metric fetcher.
- [Data Exploration](https://confidentprep.com/paths/ml-engineering-on-aws/data-exploration/): An EDA pipeline for the UCI Adult Income dataset that quantifies class imbalance, computes per-feature statistics, flags high-missing-rate columns for imputation, and surfaces correlated features in a structured Markdown data quality report.
- [Feature Engineering](https://confidentprep.com/paths/ml-engineering-on-aws/feature-engineering/): Reusable sklearn preprocessing pipeline with before/after validation — the same script runs locally and as a SageMaker Processing Job.
- [Manual Versioning](https://confidentprep.com/paths/ml-engineering-on-aws/manual-versioning/): Train 3 XGBoost variants and track results with timestamped folders and JSON metadata, before MLflow takes over.
- [Experiment Tracking](https://confidentprep.com/paths/ml-engineering-on-aws/experiment-tracking/): MLflow-instrumented XGBoost training with a hyperparameter sweep, programmatic best-run selection, model registry with a production alias, and inference loaded by that alias.
- [Batch Inference](https://confidentprep.com/paths/ml-engineering-on-aws/batch-inference/): Packages an MLflow-registered model for SageMaker Batch Transform, runs offline predictions on a held-out test set, and compares batch vs real-time endpoint cost.
- [Model Serving](https://confidentprep.com/paths/ml-engineering-on-aws/model-serving/): Deploy an MLflow-registered model as a real-time SageMaker endpoint, benchmark p50/p95/p99 latency, and measure cold start time.
- [A/B Testing](https://confidentprep.com/paths/ml-engineering-on-aws/ab-testing/): Deploy two model variants behind one SageMaker multi-variant endpoint, shift traffic, and promote the winner by AUC.
- [Performance Reporting](https://confidentprep.com/paths/ml-engineering-on-aws/performance-reporting/): Generate a weekly Markdown report comparing live endpoint AUC and CloudWatch health to the MLflow registry baseline, safe to run as a cron job.
- [Model Monitoring](https://confidentprep.com/paths/ml-engineering-on-aws/model-monitoring/): Set up SageMaker Model Monitor for data drift detection: data capture, baseline statistics, hourly schedule, synthetic drift injection, plain-English violation reports.
- [Retraining Pipeline](https://confidentprep.com/paths/ml-engineering-on-aws/retraining-pipeline/): Wire a Step Functions state machine triggered by S3 uploads: preprocess, train, evaluate against an SSM baseline, and conditionally deploy with SNS notifications.
- [MLOps Capstone](https://confidentprep.com/paths/ml-engineering-on-aws/capstone-mlops/): Wire model serving, A/B testing, drift monitoring, and automated retraining (p3-07 through p3-11) into one observable, documented MLOps system.

## AI/ML Interview Course

- [Session 1: How LLMs Actually Work](https://confidentprep.com/courses/ai-ml-for-interview/1-how-llms-actually-work/)
- [Session 2: Prompt Engineering & the LLM API Surface](https://confidentprep.com/courses/ai-ml-for-interview/2-prompt-engineering-and-the-llm-api-surface/)
- [Session 3: Embeddings & RAG](https://confidentprep.com/courses/ai-ml-for-interview/3-embeddings-and-rag/)
- [Session 4: Context & Memory Management](https://confidentprep.com/courses/ai-ml-for-interview/4-context-and-memory-management/)
- [Session 5: Tool Use & Single-Agent Loops](https://confidentprep.com/courses/ai-ml-for-interview/5-tool-use-and-single-agent-loops/)
- [Session 6: Multi-Agent Systems & Role-Play](https://confidentprep.com/courses/ai-ml-for-interview/6-multi-agent-systems-and-role-play/)
- [Session 7: Guardrails, Evals & Observability for Agentic Systems](https://confidentprep.com/courses/ai-ml-for-interview/7-guardrails-evals-and-observability-for-agentic-systems/)
- [Session 8: Model Selection & Fine-Tuning Trade-offs](https://confidentprep.com/courses/ai-ml-for-interview/8-model-selection-and-fine-tuning-trade-offs/)
- [Session 9: Production AI System Design](https://confidentprep.com/courses/ai-ml-for-interview/9-production-ai-system-design/)
- [Session 10: AI Security & Responsible AI](https://confidentprep.com/courses/ai-ml-for-interview/10-ai-security-and-responsible-ai/)
- [Session 11: Capstone Build](https://confidentprep.com/courses/ai-ml-for-interview/11-capstone-build/)
- [Session 12: Mock Interview & Portfolio Narrative](https://confidentprep.com/courses/ai-ml-for-interview/12-mock-interview-and-portfolio-narrative/)

## Articles

- [The LLM Fundamentals Interview Question That Trips Up Senior Candidates](https://confidentprep.com/blog/1-llm-fundamentals-interview-questions-explained/): "How does an LLM generate a response?" sounds like a warm-up question. It isn't. Here's why most answers stall after two beats, and what the full answer actually requires. Tags: interview-prep, llm, career, fundamentals.
- [What Your LLM Prompts Actually Cost (Most Teams Never Measure This)](https://confidentprep.com/blog/1-what-your-llm-prompts-actually-cost/): Token count, not word count, is what you're billed for — and the gap between the two is bigger than most engineers assume. A framework for measuring prompt cost before it shows up as a surprise on the invoice. Tags: llm, cost, engineering, fundamentals.
- [What Separates a Senior Prompt-Engineering Answer From a Junior One](https://confidentprep.com/blog/2-prompt-engineering-interview-questions-explained/): Interviewers ask one open-ended question about production prompt design and listen for five specific things. Here's what those five things actually are, and which one most candidates skip. Tags: interview-prep, prompt-engineering, career.
- [The Four Fields Every Production Prompt Needs (and What Breaks When You Skip One)](https://confidentprep.com/blog/2-the-rctf-framework-for-production-prompts/): Role, Context, Task, Format — the RCTF framework for writing prompts that hold up in production, plus the measured cost of skipping the Format field: an 80% format-drift rate. Tags: prompt-engineering, llm, ai-engineering.
- [What Agent-Loop Interview Questions Are Actually Testing For](https://confidentprep.com/blog/5-agent-loop-interview-questions-explained/): Interview questions about tool use and agent loops aren't really testing whether you can name the steps — they're testing whether you've watched one fail. Here's what the questions are really probing for. Tags: ai-agents, interview-prep, tool-use.
- [The Permission Model Hiding Inside Every 'AI Agent'](https://confidentprep.com/blog/5-how-ai-agents-actually-call-tools/): Every AI agent that takes real actions relies on a boundary most people never name explicitly: the model requests, your code decides. Here's why that boundary is the whole design, and what breaks when it's missing. Tags: ai-agents, tool-use, system-design, python.
- [Multi-Agent AI Systems: The Two Bugs That Only Show Up After You Add a Second Agent](https://confidentprep.com/blog/6-designing-a-multi-agent-ai-team/): A single AI agent fails loudly. A team of them can fail silently — passing garbage from one agent to the next while reporting success. Here's the design pattern that catches both failure modes before production does. Tags: ai, agents, system-design, engineering.
- [What Interviewers Are Actually Testing When They Ask About Multi-Agent Systems](https://confidentprep.com/blog/6-multi-agent-interview-questions-explained/): Multi-agent system design has become one of the default AI engineering interview topics. The question isn't whether you can name the pattern — it's whether you can explain what breaks it. Tags: ai, interview-prep, agents, career.
- [What Interviewers Are Actually Listening For When They Ask About AI Safety](https://confidentprep.com/blog/7-guardrails-evals-interview-questions-explained/): Production AI safety interview questions have a shape: six specific beats, and most candidates only cover four of them. Here's what the missing two reveal about your answer. Tags: ai, interview-prep, agents, production.
- [The Three Layers Between a Demo and a Production AI Agent](https://confidentprep.com/blog/7-the-production-ai-safety-stack/): Guardrails, evals, and observability aren't three separate projects — they're three layers of the same safety stack, and most teams only build the first one. Here's why that's not enough. Tags: ai, production, agents, safety.
- [Choosing the Right Model Tier: The Cost-Latency-Quality Triangle](https://confidentprep.com/blog/8-choosing-the-right-model-tier/): Most model-selection debates start with a model name. The ones that hold up under scrutiny start with a constraint. Here's the framework, with real cost numbers. Tags: ai, model-selection, architecture.
- [The Model-Selection Interview Question Almost Everyone Answers Backwards](https://confidentprep.com/blog/8-model-selection-interview-questions-explained/): Interviewers don't want a model name. They want to hear you name a constraint first. Here's why that one habit is the whole signal — and what it reveals about fine-tuning, benchmarks, and production trade-offs. Tags: ai, interview-prep, model-selection.
- [AI Portfolio Projects for Career Switchers: What to Build When Your Past Experience Is Not AI](https://confidentprep.com/blog/ai-portfolio-projects-for-career-switchers/): A career-switcher guide to AI portfolio projects: how engineers, data analysts, PMs, and ML-adjacent professionals can choose projects that make past experience relevant. Tags: portfolio, career-change, projects.
- [Best AI/ML Projects for Your Resume: 12 Portfolio Builds That Prove Real Skill](https://confidentprep.com/blog/best-ai-ml-projects-for-resume/): A practical AI/ML resume project list: RAG, agents, evals, streaming APIs, SageMaker, monitoring, and other portfolio builds hiring teams can inspect. Tags: portfolio, career, projects.
- [How to Build an AI Portfolio With No Experience: Start Small, Then Prove Production Skill](https://confidentprep.com/blog/how-to-build-ai-portfolio-no-experience/): A practical roadmap for building an AI portfolio from zero experience: start with small LLM tools, add RAG, evals, deployment, and a public project narrative. Tags: portfolio, beginner, career.
- [Why Your LLM App 'Forgets' — and the Three Memory Strategies That Fix It](https://confidentprep.com/blog/llm-context-and-memory-management/): An LLM has no memory of its own — every call only sees what you send it. Here's why a bigger context window doesn't solve that, and the three strategies that actually do. Tags: llm, memory, architecture.
- [Deploying ML Models on AWS SageMaker: From Notebook to Endpoint](https://confidentprep.com/blog/deploying-ml-models-aws-sagemaker/): Running a model in a Jupyter notebook proves nothing. Here's how to deploy a real SageMaker endpoint with autoscaling — using boto3 and the SageMaker SDK. Tags: aws, sagemaker, ml, python, inference.
- [Fine-Tuning vs Prompt Engineering: How to Pick the Right Lever](https://confidentprep.com/blog/fine-tuning-vs-prompt-engineering/): Most teams jump to fine-tuning too early. Here are the 3 questions to ask first, when prompting wins, and a Python benchmark to measure the difference yourself. Tags: ml, fine-tuning, python, learning.
- [Understanding RAG: When to Use It and When Not To](https://confidentprep.com/blog/understanding-rag-architecture/): Retrieval-Augmented Generation is everywhere. But most tutorials skip the tradeoffs. Here's what actually matters when you're building production RAG. Tags: rag, architecture, aws.
- [AWS IAM for ML Engineers: Least-Privilege Roles That Actually Work](https://confidentprep.com/blog/aws-iam-for-ml-engineers/): ML teams consistently over-permission their AWS roles. Here's how execution roles actually work, what permissions SageMaker jobs actually need, and a Terraform module to create it right. Tags: aws, iam, ml, sagemaker.
- [Why Projects Beat Courses for AI/ML Learning](https://confidentprep.com/blog/why-projects-beat-courses/): Certificates don't get you hired. Deployed projects do. Here's the difference between passive and active learning in AI. Tags: learning, career.

## Machine-Readable Files

- [llms.txt](https://confidentprep.com/llms.txt)
- [llms-full.txt](https://confidentprep.com/llms-full.txt)
- [RSS feed](https://confidentprep.com/rss.xml)
- [Sitemap index](https://confidentprep.com/sitemap-index.xml)

## Source

- [GitHub — ai-engineering-projects](https://github.com/confidentprep14-web/ai-engineering-projects): open-source project implementations for the paths above.