The Venetia Project began with a gap.
H. H. Asquith wrote hundreds of letters to Venetia Stanley—letters that survive in remarkable detail. But Venetia's own voice during these years is largely missing. We hear him constantly: his doubts, his political calculations, his emotional dependence. What we don't have is a clean, balanced record of what surrounded those letters day by day.
At some point, that imbalance stopped feeling like a limitation and started feeling like a challenge.
So instead of trying to "complete" Venetia's voice, I decided to do something else.
I decided to approach the problem the way I usually do when something feels unresolved: by experimenting.
I'm a technologist, and I'm deeply interested in what recent AI tools make newly possible—not in replacing human judgment, but in handling scale, complexity, and cross-reference in ways that were previously impractical. This felt like exactly the kind of historical problem those tools might help with.
So instead of trying to reconstruct Venetia's missing voice directly, I decided to feed my AI everything else.
Every letter I could find.
Diaries of people around them.
Cabinet minutes. Parliamentary debates. Newspapers. Political crises. Travel records. Social events. Even the weather.
Once the data was ingested, I started asking questions:
The experiment was simple in spirit but ambitious in scale: to see whether modern AI tools could help reconstruct the daily texture of a historical relationship without smoothing it into a story it never was.
The Venetia Project is designed to be explored, not consumed linearly.
It includes:
Some days are dense. Some are empty. Both matter.
This project isn't finished, and it isn't definitive.
If you spot an error, disagree with an interpretation, have a question, or know of a source I should look at—I'd genuinely like to hear from you.
You can reach me at: elon@consi.io
History gets more interesting when it's examined closely, and more honest when its gaps are left visible.
This project was built through multiple AI tools working together, each chosen for its strengths.
Served as the project's "brain," ingesting historical documents to reason across raw data without the bias of hindsight.
Functioned as the chief assistant for infrastructural work: writing Python code to download, scrape, and reconcile data from various sources.
Configured as a history assistant to find additional primary sources by answering targeted questions about dates, authors, and corroborating materials.
Acted as a "thinking partner" for reframing questions and resolving narrative difficulties when data models hit a wall.
Collaborated with Claude to stabilize the UI produced by Base44, transitioning exploratory designs into a functional platform.
My main development environment—connecting AI-assisted reasoning directly to the codebase while building the site.
To generate the audio reconstruction of Asquith reading his letters.
Underpins the archive with Vector Search, storing embeddings to facilitate live, conversational interaction across thousands of documents.
Used heavily for "vibe-coding": enabling rapid iteration on layout, structure, and UI to maintain flexibility as ideas evolved.
Together, these tools made it possible to build a research system that combines storytelling, data, and historical methodology.
Contemporary letters written by the Prime Minister. They offer a unique, unfiltered view of the executive mindset, containing state secrets and private anxieties shared in real-time.
Raw transcripts of private letters. Highly credible evidence of the personal relationship and social maneuvering between Venetia and Edwin.
Valuable for access to the PM's inner circle, though editors warn Margot was "an opinionated egotist, often inaccurate... and occasionally prone to fantasy."
Edited selections of diaries and letters from Asquith's daughter, intensely loyal to the Asquithian liberal viewpoint.
An "intimate, unselfconscious record." Highly reliable for social history and the mood of the aristocracy during the war.
Stevenson was both Lloyd George's secretary and mistress. The diary reflects his biases and justifications from the center of power.
Riddell was a press baron and intermediary. Highly reliable regarding the relationship between the media and the government.
Written in 1958, this is a retrospective memoir. It captures the spirit and romance of the Coterie and the era.
Hankey was Secretary to the War Council. Considered a definitive administrative history of the war's direction.
Verbatim transcripts of speeches in the House of Commons. The definitive source for what was publicly said in Parliament.
Raw primary documents: telegrams, secret memos, and private letters from Churchill, Asquith, Fisher, and others.
Historical weather data used to contextualize daily events and confirm atmospheric details mentioned in letters.
Contemporaneous accounts of public events, political developments, and social context from The Times, Liverpool Echo, Yorkshire Post, etc.
Alongside these, I deliberately use secondary sources—biographies and historical studies—not to settle debates, but to show how historians have interpreted the same evidence over time. Including those perspectives makes disagreements, assumptions, and blind spots visible rather than implicit. This is not about replacing historians, but about placing their interpretations back next to the raw material.
A seminal biography written by a former Home Secretary and Chancellor. It offers deep political insight into Asquith's career and decisions.
Published in 2002, this is a synthesis of diaries and letters. It provides a reliable narrative overview, utilizing primary sources.
This text reconstructs the life of Edwin Montagu using his letters and other primary archives.