Content Taxonomy and Schema for Turing Post Newsletter

Date: 2025-12-01 Purpose: Define a consistent frontmatter schema for all newsletter content Source: Analysis of ~130 emails in content/uncategorized/

Executive Summary #

The Turing Post newsletter contains several distinct content series, each with identifiable patterns in titles. This document defines a frontmatter schema that can be applied consistently across all emails to enable filtering, categorization, and content discovery.

Content Series Identification #

Primary Series (Numbered Episodes) #

Non-Numbered Series #

Non-Series Content Types #

Frontmatter Schema #

Required Fields #

1---
2title: "string"           # Cleaned title (remove redundant prefixes if needed)
3date: YYYY-MM-DD          # Publication date
4series: "string|null"     # One of: fod, topic, agentic, insights, concepts, podcast, unicorn, guestpost, or null
5content_type: "string"    # One of: article, digest, explainer, interview, profile, curated, announcement, sponsored, forwarded, survey
6---

Series-Specific Fields #

1---
2episode: integer          # Episode number when part of numbered series (e.g., 64 for FOD#64)
3---

Categorization Fields #

1---
2primary_topic: "string"   # Main subject area (see Topic Categories below)
3tags: ["string", ...]     # Specific technologies, techniques, or concepts
4---

People and Organizations #

 1---
 2people_mentioned:         # Notable individuals referenced in the content
 3  - name: "string"        # Full name
 4    role: "string"        # Their role/title (optional)
 5    affiliation: "string" # Company/institution (optional)
 6
 7companies_mentioned:      # Organizations referenced
 8  - "string"              # Company name
 9
10models_mentioned:         # AI models referenced
11  - "string"              # Model name (e.g., "GPT-4", "Llama 3", "DeepSeek-V3")
12---

Content Characteristics #

1---
2audience: "string"        # One of: technical, general, business, mixed
3depth: "string"           # One of: overview, intermediate, deep-dive
4has_research_papers: bool # Contains academic paper references
5has_code_examples: bool   # Contains code snippets
6is_premium: bool          # Premium subscriber content
7---

Source Metadata #

1---
2source: "Turing Post"     # Always "Turing Post" for this newsletter
3author: "string"          # Primary author (default: "Ksenia Se")
4original_subject: "string" # Raw email subject line
5original_file: "string"   # Original .eml filename
6url: "string"             # Canonical URL (from footer: turingpost.com/p/...)
7---

Field Value Reference #

primary_topic Values #

content_type Values #

series Values #

Example Frontmatter #

FOD Digest Example #

 1---
 2title: "Golden Age for Indie Devs and Engineers"
 3date: 2024-08-26
 4series: fod
 5episode: 64
 6content_type: digest
 7primary_topic: industry
 8tags: ["cursor", "lerobot", "indie-hacking", "robotics", "fine-tuning"]
 9people_mentioned:
10  - name: "Andrew Ng"
11    affiliation: "Landing AI"
12  - name: "Andrej Karpathy"
13    affiliation: "OpenAI (former)"
14companies_mentioned: ["Cursor", "Hugging Face", "OpenAI", "Anthropic", "Mistral", "Aleph Alpha"]
15models_mentioned: ["GPT-4o", "Pharia-1-LLM", "Jamba-1.5"]
16audience: mixed
17depth: overview
18has_research_papers: true
19has_code_examples: false
20is_premium: false
21source: "Turing Post"
22author: "Ksenia Se"
23original_subject: "FOD#64: Golden Age for Indie Devs and Engineers"
24original_file: "FOD#64_ Golden Age for Indie Devs and Engineers.eml"
25url: "https://www.turingpost.com/p/fod64"
26---

Topic Explainer Example #

 1---
 2title: "What is FSDP and YaFSDP?"
 3date: 2024-06-20
 4series: topic
 5episode: 4
 6content_type: explainer
 7primary_topic: infrastructure
 8tags: ["fsdp", "yafsdp", "distributed-training", "gpu-optimization", "pytorch"]
 9people_mentioned: []
10companies_mentioned: ["Meta", "Yandex"]
11models_mentioned: ["Llama 2", "Llama 3"]
12audience: technical
13depth: deep-dive
14has_research_papers: true
15has_code_examples: false
16is_premium: true
17source: "Turing Post"
18author: "Ksenia Se"
19original_subject: "Topic 4: What is FSDP and YaFSDP?"
20original_file: "Topic 4_ What is FSDP and YaFSDP_.eml"
21url: "https://www.turingpost.com/p/yafsdp"
22---

Unicorn Profile Example #

 1---
 2title: "Glean: How to Outpace Competitors in Enterprise AI"
 3date: 2024-10-12
 4series: unicorn
 5episode: null
 6content_type: profile
 7primary_topic: business
 8tags: ["enterprise-search", "rag", "knowledge-management", "startup"]
 9people_mentioned:
10  - name: "Arvind Jain"
11    role: "CEO & Co-founder"
12    affiliation: "Glean"
13  - name: "Tony Gentilcore"
14    role: "Co-founder"
15    affiliation: "Glean"
16  - name: "Bipul Sinha"
17    affiliation: "Rubrik"
18companies_mentioned: ["Glean", "OpenAI", "Google", "Rubrik", "Salesforce"]
19models_mentioned: ["BERT"]
20audience: business
21depth: deep-dive
22has_research_papers: false
23has_code_examples: false
24is_premium: true
25source: "Turing Post"
26author: "Ksenia Se"
27original_subject: "Glean: How to Outpace Competitors in Enterprise AI (and Frustrate OpenAI in the Process)"
28original_file: "Glean_ How to Outpace Competitors in Enterprise AI (and Frustrate OpenAI in the Process).eml"
29url: "https://www.turingpost.com/p/glean"
30---

Podcast Interview Example #

 1---
 2title: "Sharon Zhou on AI Hallucinations, Agents Hype, and Giving Developers the Keys to GenAI"
 3date: 2025-03-23
 4series: podcast
 5episode: 1
 6content_type: interview
 7primary_topic: llm
 8tags: ["hallucinations", "fine-tuning", "agents", "lamini", "enterprise-ai"]
 9people_mentioned:
10  - name: "Sharon Zhou"
11    role: "CEO & Co-founder"
12    affiliation: "Lamini"
13  - name: "Andrew Ng"
14    affiliation: "Stanford / DeepLearning.AI"
15  - name: "Andrej Karpathy"
16    affiliation: "OpenAI (former)"
17companies_mentioned: ["Lamini", "OpenAI", "Anthropic", "Colgate", "Meta"]
18models_mentioned: ["Llama", "DeepSeek", "GPT-4"]
19audience: mixed
20depth: intermediate
21has_research_papers: true
22has_code_examples: false
23is_premium: false
24source: "Turing Post"
25author: "Ksenia Se"
26original_subject: "🎙️🧩 TP/Inference: Sharon Zhou on AI Hallucinations, Agents Hype, and Giving Developers the Keys to GenAI"
27original_file: "🎙️🧩 TP_Inference_ Sharon Zhou on AI Hallucinations, Agents Hype, and Giving Developers the Keys to GenAI.eml"
28url: "https://www.turingpost.com/p/sharonzhou"
29---

Content to Exclude or Flag #

Should Be Excluded #

Should Be Flagged for Review #

Notable People (Frequently Mentioned) #

For reference, these individuals appear across multiple articles:

Implementation Notes #

1. Series Detection Priority: Check for emoji prefixes first (🦸🏻, 🌁, 🎙️), then text patterns (FOD#, Topic, Concepts:, Guest Post:)

2. Episode Number Extraction: Use regex to extract numbers after series identifiers

3. URL Extraction: Parse the footer text for turingpost.com/p/ pattern

4. People Extraction: Look for patterns like "Name, Role at Company" or "Name from Company" or quoted attributions

5. Model Detection: Maintain a list of known model names (GPT-*, Claude, Llama, Mistral, DeepSeek, etc.)

6. Company Detection: Look for known AI companies plus any mentioned in funding/acquisition context