1996

The Problem With "Chemistry"

In 1996, the internet was young, and science had a search problem. Type "chemistry" into AltaVista, Yahoo, or the early Google, and you'd get a flood of results pointing you to chemistry.com — a dating website. Search for "biology" and wade through pages of general noise before finding anything a working scientist or curious student could actually use.

Four colleagues running an environmental testing laboratory in Mississippi noticed this gap. Between managing lab instrumentation and doing repair work for one of the major analytical instrument manufacturers, they had a side project idea: build a search engine that was exclusively for science. No dating websites. No lifestyle content. Just real science.

"Back then, if you searched for science on the major search engines, you got a little science and a lot of garbage. We wanted to build something that didn't really exist yet."

The project was called SciSeek. It launched that year as a hand-curated science directory — every link vetted, every category organized with care. It was a labor of love from the start.


1998 – 2005

Building the Spider

A hand-curated directory has a ceiling. The web was growing faster than any team could keep up with, and relying on website owners to submit their own links meant the gaps were enormous. After a couple of years it became clear: to do this right, SciSeek needed to crawl the web itself.

So that's what happened. Research, planning, and a lot of late nights produced a custom web spider — a program that automatically finds, visits, and indexes science websites across the internet, much like Google does for the whole web, but focused entirely on science. The index grew quickly. At its peak, SciSeek had catalogued over 20 million science-related websites.

1996
Year Founded
20M+
Sites Indexed
2M+
Searches Per Day

At the height of its traffic, SciSeek was serving over 2 million science searches every day. Scientists, students, researchers, and the simply curious were finding real science answers in a place built specifically for them.

There was even a highlight moment that felt like validation: NASA came on board as an advertiser, funding a campaign that redesigned the front page with a rocket shooting across the screen on every visit. It was the kind of partnership that felt like proof the idea was working.


2005 – 2008

The Economics of Running a Search Engine

Running a search engine at scale is expensive. Servers, bandwidth, storage for tens of millions of indexed pages — the costs mounted faster than the revenue could keep up. Monetization was tried from every angle: banner advertising, affiliate programs (few science-relevant ones existed at the time), science magazine subscriptions, science news feeds, and more. None of it generated enough to justify the infrastructure.

The web spider and its massive index eventually became unsustainable to maintain. Hosting the servers and managing all that data simply cost too much. In 2008, the team made the difficult decision to officially wind down the science search engine and the active development behind it.

"We officially stopped trying around 2008 — but I just couldn't let it go."


2008 – Today

Kept Alive by Stubbornness and Hope

Most side projects that stop working get abandoned. SciSeek didn't. For over 15 years after winding down active development, the site kept running in a quieter form — powered by Google's Programmable Search Engine, maintained by one person still convinced the original idea was worth something, still occasionally tinkering, still dreaming about what it could become.

That person is still here. Still a field service engineer working with the analytical instrumentation that's been a thread running through this whole story. Still traveling, still working on SciSeek in the downtime between jobs, in hotel rooms and airports, because some ideas are just too good to give up on.

When Google announced it was ending its Programmable Search Engine service in January 2027, it forced a decision: let SciSeek finally go, or find a new way forward. The answer, as it turns out, was obvious.


Now

Science Questions, Finally Answered

The original promise of SciSeek was simple: give people a place to find real science, without the noise. In 1996, doing that required building a directory, then a spider, then a 20-million-page index. In 2025, it turns out the answer fits in a single text box.

The new SciSeek is powered by AI — capable of answering science questions across every discipline, from quantum physics to evolutionary biology, from chemistry to cosmology. No garbage results. No dating websites. Just science.

The mission hasn't changed in nearly 30 years. Only the technology has.

Welcome to SciSeek Version 2.0!