We build AI that helps people decide with evidence
Abbadh Labs is an AI deep-tech company commited to transforming fragmented data into clear, actionable insights through advanced autonomous research, predictive analytics, and decison-intelligence systems.
Every day, universities, hospitals and institutions collect data that could change a decision. Most of it never does. It sits in folders, spreadsheets and filing cabinets while the same questions get answered by guesswork.
We started Abbadh Labs in 2021 because we believe whoever understands their data makes better decisions. Our mission is to give people that understanding: systems that read the evidence, find the pattern and show what it means, so the person deciding can act on knowledge instead of assumptions.
We began with research, because research is where evidence is made, and because students were being let down by tools that wrote confidently and cited things that didn't exist. We are bootstrapped, funded by our own capital, family support and product revenue, and answer only to the people who use what we build.
For years, researchers didn't trust AI with their work. They were right not to.
The tools made up references, ignored academic structure, and stopped long before a chapter was finished. We studied each of those failures and built SABABAT specifically to fix them. Today more than 20,000 researchers use it.
Every citation is a paper we actually found
SABABAT searches across its entire database of over 250 million academic and scientific research, before it writes a word. Weak and off-topic papers are filtered out, and it never falls back on references a model "remembers".
Chapters built the way examiners expect
Background to scope in Chapter One; conceptual, theoretical and empirical review in Chapter Two. Paste your preferred outline into the Authorial Anchor and SABABAT follows it.
Complete chapters, at the depth your level demands
A Chapter Two runs to 6000+ words with up to 25 or more verified references, or 6,000+ words and 32 references for postgraduates, with tables, framework diagrams and graphs if you want them. Chapters One to Five, essays and full proposals.
An assistant that will tell you no
If a research design can't answer its own objectives, SABABAT flags it, choose and explains the right approach.
AI should make research faster to do, never easier to fake.
Fluency is cheap. Evidence isn't.A model can write a convincing paragraph about anything. That's exactly why we make our systems find the source first and write second. A claim with no source gets no citation.
Statistics is a method, not decoration.The right test comes from the question, the design and the data, not from what looks impressive in a results table. A significant p-value from the wrong design is still the wrong answer.
The researcher stays the author.People bring the topic, the objectives and the argument. Our job is to support and challenge that thinking, and to leave every final judgment with the person who has to stand behind it.
The people behind
Abbadh Labs
Help us build what comes next
We're a small team, and everyone works on real product decisions with real users. There's no HR department: email us what you would build, fix or improve. That's the application.
Get in touch →




