Postdoc website examples: six real sites, annotated

A postdoc's website is the application packet, in public. It is what a search committee or a prospective PI opens before the PDF, and it is the one address that does not expire with the contract. This page shows six real postdoc websites — all live, all linked, none of them ours — each reviewed critically: what it does well, what is missing, and what is worth stealing. Then it answers a question nobody else on this topic has measured: what actually happens to postdoc websites after a few years. This is not a list of postdoc job boards. For sites across every career stage, see our twelve academic website examples across every career stage.

All six sites were visited and reviewed on October 13, 2026, and roles and affiliations are as stated on each site that day. None of these researchers is affiliated with Scilio, and several are hosted on platforms that compete with us. That is the point. Sites change. If an annotation no longer matches, tell us: support@scilio.app.

What a postdoc's website has to do (that a PhD student's and a professor's don't)

What a PhD student's website has to do is make a thin record look like a direction. A professor's site has to serve three audiences at once, which is why we keep professor website examples, annotated through a professor's lens separately. A postdoc's site has a different shape, in three ways that no ranking page addresses.

It is the application packet, in public. Postdocs are on the market more or less continuously. The site is what gets opened before the PDF, and it is the only version of your record that you control the order of. One of the six sites below links its CV, research statement and teaching statement from the homepage. That costs a little privacy and buys a lot of time for the reader who matters.

It has to outlive the address. Postdocs move more than anyone else in academia, and the institutional profile dies with each contract. In Nature's 2023 global postdoc survey, 61% of the 3,838 respondents were working outside their home country, and career dissatisfaction was higher among people on their third or fourth postdoc (34%) than on their first (24%). The 2020 edition of the same survey, with 7,670 respondents, put the share working outside their home country at 61.4%. Both surveys were self-selecting and advertised through Nature's own channels, so read them as a picture of who answered, not a census. Neither says a word about personal websites. They establish only that postdocs move, which is the whole argument for owning the address rather than renting it from a department.

It will become something else. The postdoc site is the one most likely to be rebuilt: into a lab site after a faculty hire, into an industry page, or into nothing. Our own check of a 2019 cohort of 25 postdoc sites, below, puts numbers on that. Two of the six examples show what a site looks like when it was built to grow.

Six real postdoc websites, annotated

Each of these was visited and reviewed on October 13, 2026. Two of them also appear in our broader examples collection with different notes. Here they are read as postdoc sites.

Nicole Paul — Disaster risk and earthquake engineering, ETH Zürich

Hugo Blox · nicolepaul.io

The header says “ETH Fellow” and the first line of the bio says, in her own words, “I've recently joined ETH Zürich as a postdoctoral research fellow”. A fellowship is a postdoc with a name, and the site treats it that way: a dated experience block (Arup 2014–2019, Global Earthquake Model 2019–2023, PhD at UCL 2026, ETH 2026–present) that makes a decade of practitioner work legible to a committee in one screen. Featured publications carry PDF, Cite, Dataset, Project, Code and DOI rows where each exists, and the newest entry is dated May 2026 in the International Journal of Disaster Risk Reduction. The source says “Last Published: May 21, 2026” and the footer reads “© 2023–2026”.

What's missing: The posts stopped. The most recent one is “Tools I've found useful during my PhD”, last updated July 29, 2024, under a bio that has since moved countries and career stages. Three research topics are listed with the supervisor who oversaw each, which reads as a student's framing on a site that is no longer a student's.

What to steal: Let the right thing slip. The publication list stayed current while the blog did not, and for a postdoc that is the correct trade. Nobody hires on blog cadence.

Alexandra de Gendre — Economics, University of Melbourne and CNRS

Quarto · adegendre.github.io

Two continents, one URL. She is the “2022 McKenzie Postdoctoral Research Fellow” at Melbourne since February 2022 and a CNRS Research Fellow at CREST in Paris since February 2026, after a PhD at Maastricht in 2021 and a first postdoc in Sydney. The institutional pages for those four stops are four different websites. Hers is one. The research page is the economist's answer to “under review”: every paper carries its status in brackets — “Conditionally Accepted at the Review of Economics and Statistics”, “Revise and Resubmit”, “Under Review”, “Accepted” — with the working-paper PDF, the IZA or CEPR discussion-paper version, slides, media coverage and replication data linked beneath it.

What to steal: Label the status of every paper. “Under review” with a PDF is a result. A title with nothing under it is a rumour.

Nathan Constantine-Cooke — Health data science, University of Edinburgh

Jekyll on a custom domain · constantine-cooke.com

“Postdoctoral Research Associate in Health Data Science” at the Institute of Genetics and Cancer, and the site has the one section almost no postdoc builds: Repositories. It lists his R packages by name (datefixR under rOpenSci, gameR, bibget, two group packages) with GitHub stats beside them. For a methods postdoc that page is a stronger exhibit than the publication list, and the publication list is good too: journal abbreviations, asterisks marking co-first authors, Abs and Bib buttons that expand the abstract and the BibTeX entry in place. The footer says 2026.

What's missing: The news feed froze at “I passed my viva!” on January 31, 2024, under a heading that still calls the site new (“This website has gone live!”, February 2022). The Teaching and Projects pages exist because the template had them. A reader learns that the site was built with care as a PhD student and has been coasting since.

What to steal: A Repositories section. If your work ships as code, show the code where a committee can count the downloads. And the lesson to avoid: a dated news feed is a liability the day you stop feeding it.

Inho Choi — Computer systems, ETH Zurich

Hugo, Eureka theme · ihchoi12.github.io

The application packet in public, literally. Under the name and “Postdoctoral Researcher, ETH Zurich, Department of Computer Science” sit three buttons: CV, Research Statement, Teaching Statement. Then Scholar, LinkedIn and an email address. The publications are venue-tagged in the systems convention — “[SIGCOMM '26] Capybara: Dynamic Load Balancing with Microsecond-Scale TCP Migration” — and several expand into a plain-language summary of what the project did and what his part of it was. The experience block tells the postdoc story without comment: PhD at the National University of Singapore to January 2026, a postdoc there from February to September 2026, ETH from October 2026.

What's missing: “© 2022 Inho Choi”. Two postdocs and a continent later, the footer still says the year the template was installed. The site has no narrative beyond the research interests paragraph, and the statements it links are PDFs, so the web page itself says less than the documents behind it.

What to steal: Put the statements on the homepage. If a committee is going to ask for them anyway, the version they open first should be the one you ordered. And fix the footer year. It takes a minute and it is the first thing a careful reader notices.

Himanshu Beniwal — Multilingual NLP and AI safety, TU Dresden

Jekyll on GitHub Pages · himanshubeniwal.github.io

Current to the month, which almost nobody manages. “Postdoctoral Researcher, ScaDS.AI / Technische Universität Dresden” since July 2026, a PhD at IIT Gandhinagar, a Microsoft Research India internship from February to May 2026, and a news feed whose top item is dated October 2026 with fifty earlier items folded away. The footer says “Last updated: August 24, 2026”. Scholar, LinkedIn, GitHub, X, Hugging Face and Instagram are all one click from the header.

What's missing: Everything is here, and that is the problem. The side navigation lists thirteen sections on one page — About, News, Experience, Research Interests, Publications, Live Projects, Talks and Posters, Education, Teaching, Service, Media Coverage, Notebooks, Past Projects — plus a posters gallery and a “Memories” page. A row of counters announces 22 publications and 320+ citations before the reader has learned what the research is. The three things a committee needs are on the page. They are just not easy to find among the other ten.

What to steal: The update discipline. A dated “last updated” line in the footer and a news item from this month do more for credibility than any design choice. Then cut: a postdoc site should have five sections, not fifteen.

Chenghao Liu — Chemistry and machine learning, Caltech

Jekyll, single page · pchliu.github.io

The two-audience site that chose. The page title is “Postdoc at Caltech”, the header holds exactly two things, a Google Scholar link and an email address, and the bio covers a FutureHouse postdoctoral fellowship with Frances Arnold (2025–present), a PhD at McGill (2019–2024) with concurrent work at Mila, a co-founded protein-design startup (2021–2025) and a Vanier scholarship, in four sentences. Then publications by year, 2026 back to 2019, with a thumbnail each, a legend for † equal contribution and * corresponding author, and a DOI or arXiv link on every entry. He also lists the workshop series he organises at ICLR, with the next edition and the competition it runs.

What's missing: No CV, no teaching, no news, no statement of what comes next. Academia and a startup sit on one page and nothing on it says which one he wants. For a postdoc with offers to choose between, that may be deliberate. For a reader deciding whether to make one, it is a gap.

What to steal: The header. Two links, the right two. And the author-contribution legend, which costs one line and answers the question every committee asks about a multi-author paper.

What the good ones have in common

Six sites, five platforms, five countries, and the same five habits wherever the site works.

1.
Papers carry their status. De Gendre labels every working paper with where it stands in review. Liu marks equal contribution and corresponding authorship. Constantine-Cooke marks co-first authors with an asterisk. A postdoc's list is mid-flight by definition, and the sites that admit it read as more credible, not less.
2.
The strongest exhibit is not always a paper. Constantine-Cooke's Repositories page and Paul's Dataset and Code rows show the work in a form a reader can run. If your field ships code, data or instruments, the publication list is the index, not the exhibit.
3.
The packet is one click away. Choi links CV, research statement and teaching statement from the homepage. Paul and de Gendre link a CV. A committee that has to email you for a document has already formed an opinion.
4.
The footer year is a tell. Choi's says 2022. De Gendre's says 2024. Paul's, Beniwal's and Constantine-Cooke's say 2026. It is the cheapest fix on any site and the one readers use, fairly or not, to decide whether the rest is current.
5.
Current to the month beats complete. Beniwal's October 2026 news item and dated “last updated” line do more than his thirteen sections. One dated line that is true this month outranks ten sections that were true last year.

For the planning walkthrough behind these habits, we keep how to make a personal academic website, step by step.

What happens to postdoc websites

Nobody on this topic has measured it, so we did. In May 2019, Jack Cao published a census of the personal websites of Harvard Psychology's graduate students and postdocs, with a line that still holds: “if these websites are that buried and hard to find, they may as well not exist.” It listed 25 postdocs with a website, each with its URL. On October 13, 2026, seven and a half years later, we requested all 25.

Result, October 13, 2026	Sites
Still loads	15
Confirmed gone (410, 404, or the domain no longer resolves)	5
Institutional host that refuses automated requests, so unknown	5

Three of the five that are gone were custom domains that nobody renewed. One was a page on a university blog platform that now returns “410 Gone”. One was a GitHub Pages site that now says “Site not found”. The five we could not check all sit on one institutional profile host, and we count them as neither alive nor dead.

The fifteen that load are the more interesting group, because several are no longer postdoc sites. Two now title themselves as labs, which is the faculty outcome: the personal site became the group site. Two redirect from their 2019 address to a new custom domain, which is the move outcome: the owner took the site with them and left a forwarding address behind. The rest are still a person's own page, on a mix of free hosts and paid domains.

So the honest answer to “where should I build it” is the one that survives all three outcomes. A free host under your own name survives the move. A domain you own survives the host. Either survives the department. This is the honest place to mention what we build: Scilio publishes a portfolio for free at your own scilio.app address, and a custom domain comes with the paid plan, so the address can follow you from this contract to the next. It is held to the same standard as every site above. Scilio is portfolio-only, with no lab or group sites, so the “became a lab site” outcome is one you would rebuild elsewhere, and there is no ORCID import. If you would rather fork a template and host it yourself, our academic website templates, from forkable code to live demos page is the honest comparison.

One you can click through

Screenshots, ours included, cannot show what a site feels like to navigate. So here is a live one: Dr. Ilaria Bentes — marine ecology, postdoc, a postdoctoral marine ecologist in Lisbon with heatwave-recovery research, field projects and citizen science, in the Marine theme. Ilaria is a fictional persona and her photo is AI-generated. She exists so you can click around a finished postdoc-stage site without us parading a real customer. She is built on Scilio (that's us). Her page wears the Marine theme, one of twenty a portfolio can be dressed in.

Build your own

Every postdoc above is between addresses, and every one of them has a site that will still be there after the next move. Build and publish yours for free at your own scilio.app address, publications, code and statements included, and take it with you.