Observatory · A working recordSignal · Live
What we publishwhen it stops beingload-bearing.
001 convenings · 100 insightsUpdated when files close
The Observatory is what remains when an engagement closes. Briefings, insights and instruments — released the moment they are no longer holding a live file together. No promotion, no schedule, no editorial calendar.
03 / 06 · Observatory00 scheduled · Next file
01 / What lives here
Four channels.One discipline.
- § 01
Convenings
Briefings, roundtables and closed sessions. Held under Chatham House by default.
- § 02
Insights
Working notes from active files. Released when the underlying matter is closed.
- § 03
Mandates
Public records of engagements where the client has chosen disclosure.
- § 04
Tools
Calculators, frameworks and instruments built for our own files first.
/01 · CONVENINGS
000 · SCHEDULEDBriefings & closed sessions
WindowMode
No convenings match this filter.
/02 · ARCHIVE
100 · ENTRIESInsights
FEATURED14 MAY 2026
The Next AI Bottleneck Isn’t the Model: It’s the Inference System
Medium
14 MAY 2026Medium
The Counterintuitive Networking Decisions Behind OpenAI’s 131,000-GPU Training Fabric
14 MAY 2026Medium
I Let CodeSpeak Take Over My Repository
14 MAY 2026Medium
How to Write Robust Code with Claude Code
13 MAY 2026Medium
I Built the Same B2B Document Extractor Twice: Rules vs. LLM
13 MAY 2026Medium
Exploring Patterns of Survival from the Titanic Dataset
13 MAY 2026Medium
What’s the Best Way to Brainwash an LLM?
13 MAY 2026Medium
Building an Evaluation Harness for Production AI Agents: A 12-Metric Framework From 100+ Deployments
12 MAY 2026Medium
From Vibe Coding to Spec-Driven Development
12 MAY 2026Medium
Hybrid Search and Re-Ranking in Production RAG
12 MAY 2026Medium
Proxy-Pointer Framework for Structure-Aware Enterprise Document Intelligence
12 MAY 2026Medium
Your First WebAssembly Program and Web App (Written, Tested, and Deployed Entirely in the Web Browser)
11 MAY 2026Medium
Learning Word Vectors for Sentiment Analysis: A Python Reproduction
11 MAY 2026Medium
Using Transformers to Forecast Incredibly Rare Solar Flares
11 MAY 2026Medium
How to Build a Claude Code-Powered Knowledge Base
11 MAY 2026Medium
PySpark for Beginners: Mastering the Basics
10 MAY 2026Medium
Batch or Stream? The Eternal Data Processing Dilemma
10 MAY 2026Medium
LLM Summarizers Skip the Identification Step
09 MAY 2026Medium
The Must-Know Topics for an LLM Engineer
09 MAY 2026Medium
RAG Is Blind to Time — I Built a Temporal Layer to Fix It in Production
08 MAY 2026Medium
The AI Agent Security Surface: What Gets Exposed When You Add Tools and Memory
08 MAY 2026Medium
From Data Scientist to AI Architect
08 MAY 2026Medium
When Customers Churn at Renewal: Was It the Price or the Project?
08 MAY 2026Medium
Unified Agentic Memory Across Harnesses Using Hooks
07 MAY 2026Medium
The Joy of Typing
07 MAY 2026Medium
Give Your AI Unlimited Updated Context
07 MAY 2026Medium
How Major Reasoning Models Converge to the Same “Brain” as They Model Reality Increasingly Better
07 MAY 2026Medium
I Rewrote a Real Data Workflow in Polars. Pandas Didn’t Stand a Chance.
06 MAY 2026Medium
Beyond Lists: Using Python Deque for Real-Time Sliding Windows
06 MAY 2026Medium
When the Uncertainty Is Bigger Than the Shock: Scenario Modelling for English Local Elections
06 MAY 2026Medium
Timer-XL: A Long-Context Foundation Model for Time-Series Forecasting
06 MAY 2026Medium
Why I Don’t Trust LLMs to Decide When the Weather Changed
06 MAY 2026Medium
Deconstruct Any Metric with a Few Simple ‘What’ Questions
05 MAY 2026Medium
Discrete Time-To-Event Modeling – Predicting When Something Will Happen
05 MAY 2026Medium
How to Make Claude Code Validate its own Work
05 MAY 2026Medium
RAG Hallucinates — I Built a Self-Healing Layer That Fixes It in Real Time
05 MAY 2026Medium
Surviving High Uncertainty in Logistics with MARL
04 MAY 2026Medium
Single Agent vs Multi-Agent: When to Build a Multi-Agent System
04 MAY 2026Medium
How to Build an Efficient Knowledge Base for AI Models
04 MAY 2026Medium
Playing Connect Four with Deep Q-Learning
04 MAY 2026Medium
How AI Tools Generate Technical Debt in IoT Systems — and What to Do About It
03 MAY 2026Medium
CSPNet Paper Walkthrough: Just Better, No Tradeoffs
03 MAY 2026Medium
Inference Scaling (Test-Time Compute): Why Reasoning Models Raise Your Compute Bill
02 MAY 2026Medium
Which Regularizer Should You Actually Use? Lessons from 134,400 Simulations
02 MAY 2026Medium
How a 2021 Quantization Algorithm Quietly Outperforms Its 2026 Successor
01 MAY 2026Medium
How to Get Hired in the AI Era
01 MAY 2026Medium
Churn Without Fragmentation: How a Party-Label Bug Reversed My Headline Finding
01 MAY 2026Medium
Ghost: A Database for Our Times?
01 MAY 2026Medium
Why Powerful Machine Learning Is Deceptively Easy
30 APR 2026Medium
A Gentle Introduction to Stochastic Programming
30 APR 2026Medium
Proxy-Pointer RAG: Multimodal Answers Without Multimodal Embeddings
30 APR 2026Medium
How to Study the Monotonicity and Stability of Variables in a Scoring Model using Python
30 APR 2026Medium
Why AI Engineers Are Moving Beyond LangChain to Native Agent Architectures
29 APR 2026Medium
4 YAML Files Instead of PySpark: How We Let Analysts Build Data Pipelines Without Engineers
29 APR 2026Medium
Ensembles of Ensembles of Ensembles: A Guide to Stacking
29 APR 2026Medium
Agentic AI: How to Save on Tokens
29 APR 2026Medium
System Design Series: Apache Flink from 10,000 Feet, and Building a Flink-powered Recommendation Engine
28 APR 2026Medium
Let the AI Do the Experimenting
28 APR 2026Medium
Correlation Doesn’t Mean Causation! But What Does It Mean?
28 APR 2026Medium
PyTorch NaNs Are Silent Killers — So I Built a 3ms Hook to Catch Them at the Exact Layer
28 APR 2026Medium
The Next Frontier of AI in Production Is Chaos Engineering
27 APR 2026Medium
A Career in Data Is Not Always a Straight Line, and That’s Okay
27 APR 2026Medium
How Spreadsheets Quietly Cost Supply Chains Millions
27 APR 2026Medium
Comparing Explicit Measures to Calculation Groups in Tabular Models
26 APR 2026Medium
Bytes Speak All Languages: Cross-Script Name Retrieval via Contrastive Learning
26 APR 2026Medium
I Reduced My Pandas Runtime by 95% — Here’s What I Was Doing Wrong
25 APR 2026Medium
Causal Inference Is Different in Business
25 APR 2026Medium
The Essential Guide to Effectively Summarizing Massive Documents, Part 2
24 APR 2026Medium
Introduction to Approximate Solution Methods for Reinforcement Learning
24 APR 2026Medium
I Built an AI Pipeline for Kindle Highlights
24 APR 2026Medium
How to Improve Claude Code Performance with Automated Testing
24 APR 2026Medium
How to Select Variables Robustly in a Scoring Model
23 APR 2026Medium
Using a Local LLM as a Zero-Shot Classifier
23 APR 2026Medium
I Simulated an International Supply Chain and Let OpenClaw Monitor It
23 APR 2026Medium
Your Synthetic Data Passed Every Test and Still Broke Your Model
23 APR 2026Medium
Lasso Regression: Why the Solution Lives on a Diamond
22 APR 2026Medium
Using Causal Inference to Estimate the Impact of Tube Strikes on Cycling Usage in London
22 APR 2026Medium
Correlation vs. Causation: Measuring True Impact with Propensity Score Matching
22 APR 2026Medium
From Ad Hoc Prompting to Repeatable AI Workflows with Claude Code Skills
22 APR 2026Medium
Ivory Tower Notes: The Methodology
22 APR 2026Medium
How to Run OpenClaw with Open-Source Models
21 APR 2026Medium
DIY AI & ML: Solving The Multi-Armed Bandit Problem with Thompson Sampling
21 APR 2026Medium
Git UNDO : How to Rewrite Git History with Confidence
21 APR 2026Medium
How to Call Rust from Python
21 APR 2026Medium
Your RAG Gets Confidently Wrong as Memory Grows – I Built the Memory Layer That Stops It
21 APR 2026Medium
I Replaced GPT-4 with a Local SLM and My CI/CD Pipeline Stopped Failing
20 APR 2026Medium
What Does the p-value Even Mean?
20 APR 2026Medium
Context Payload Optimization for ICL-Based Tabular Foundation Models
20 APR 2026Medium
The LLM Gamble
20 APR 2026Medium
From Risk to Asset: Designing a Practical Data Strategy That Actually Works
19 APR 2026Medium
Proxy-Pointer RAG: Structure Meets Scale at 100% Accuracy with Smarter Retrieval
19 APR 2026Medium
Dreaming in Cubes
19 APR 2026Medium
KV Cache Is Eating Your VRAM. Here’s How Google Fixed It With TurboQuant.
18 APR 2026Medium
Your RAG System Retrieves the Right Data — But Still Produces Wrong Answers. Here’s Why (and How to Fix It).
18 APR 2026Medium
AI Agents Need Their Own Desk, and Git Worktrees Give Them One
18 APR 2026Medium
Introduction to Deep Evidential Regression for Uncertainty Quantification
18 APR 2026Medium
memweave: Zero-Infra AI Agent Memory with Markdown and SQLite — No Vector Database Required
18 APR 2026Medium
Building My Own Personal AI Assistant: A Chronicle, Part 2
18 APR 2026Medium
Your Chunks Failed Your RAG in Production
18 APR 2026Medium
What It Actually Takes to Run Code on 200M€ Supercomputer
04 / Media inquiries
For journalists,editors and analysts.
We respond to media on the record only when the underlying file is closed. For background, attributable commentary, or expert introductions, write to the press desk directly.
Signal · SubscribeSignal · SubscribeSignal · SubscribeSignal · SubscribeSignal · Subscribe
05 / Signal
Receive the filethe day it closes.
A signal cut from active files. Insights, mandates and tools — published the day they stop being load-bearing for an engagement. No promotion, no tracking pixels, no schedule.
press@lequent.comNo editorial calendarLequent