About me

To get in touch, reach out to hello [at] yudhanjaya [dot] com, or contact fingan [at] zenoagency [dot] com to talk to my agent. Responses from my end may take time.
About me
Photo by Dirk Skiba Fotografie.

I write, both as a fictioneer and a journalist; I investigate systems, pasts and futures; I program; I work with, think about, and even fictionalize artificial intelligence and machine learning; and in the middle of all this I try to learn how to make things grow.

Here’s my latest bio:

Yudhanjaya Wijeratne is an author, data scientist and general tinkerer from Colombo, Sri Lanka. He is the co-founder and Editor-in-Chief of Watchdog, a research collective that works on fact checking, investigative journalism and community tech.

His fiction includes novels like Numbercaste, The Inhuman Race, The Wretched and the Damned, and The Salvage Crew. His stories have won the Gratiaen prize, been nominated for the Nebula and Independent Games Festival awards, appeared in venues like Wired, ForeignPolicy and Slate, and been Washington Post and Audible bestsellers. He has, somewhat unfortunately, also shown up on Forbes‘ 30 Under 30. He has several cats who care about none of these things, and spends most of his time on his homestead in Kandy, learning the care and feeding of potato plants.

CV, TLDR version:

  • Appointed to Sri Lanka’s National AI Advisory Committee in 2025, where I work on making it easier for citizens to interact with government via the Government Information Center, and with some ambitious dataset creation for Sinhala and Tamil - something I’ve been working around for a while now.

  • Co-founded Dossiers, a KYC + AML solution for banks, financial institutions, and journalists. Ended up on Meta’s Llama Hall of Fame, 2024, for some of the stuff we’re doing with local finetuned AI models. I build hardware and models.

  • Co-founded Watchdog, a nonprofit focused on data journalism and civic technology in Sri Lanka. As CEO, raised over $400k, built a multidisplicinary team of computer scientists, journalists and researchers, doing stuff like tracking 600+ projects, mapping mass graves using satellite imagery and old-school journalism, and building a simulation of the city of Colombo - probably the most ambitious build in Cities: Skylines.

    Led major investigations, such as using satellite imagery analysis to locate mass graves, archiving and analysis of vast troves of government data to explain electricity and infrastructure failures, and building open-source medical donations software that is subsequently being used by the Red Cross and Ministry of Health in Sri Lanka.

  • Worked at LIRNEasia, a regional think tank, working on computational social science research and machine learning. Worked on projects involving billions of call detail records, friend networks on social media, and built corpora for low-resource languages like Sinhala and Bengali. Contributed in small ways to national-level projects, including Sri Lanka’s data protection act. Published widely (specifically, corpus linguistics, social networks, misinformation and hate speech).

  • Worked at WSO2, a major middleware provider, as a writer and data science neophyte. Worked on text copy, whitepapers around client implentations (Transport For London, Govt. of Maldova, etc), and designed, prototyped and led implementation on WSO2’s election monitor.

  • Served as founding editor of Readme.LK, a leading technology publication in Sri Lanka. Grew the publication and team.

  • Started out working retail at Redline Technologies in a mall (Majestic City, Colombo). Learned a lot about how to build.

General: Forbes 30 Under 30 honoree, Gratiaen Prize winner for fiction, and Nebula and IGF award-nominated author, TEDx speaker. Largely self-taught (programming, writing and a whole lot more).

Bibliography

Novels and other major work

I’ve written a number of novels (multi-book deals with HarperCollins and Aethon), wrote for major publications, et cetera. Some things have worked out; see bio above. I write primarily in the field of SFF; it’s a vast canvas that lets me toy with ideas I find interesting. I also occasionally work with AI and software of my own design, pursuing a human-AI collaborative thesis; I’ve explored this in work with Wired, Slate and Google Research. For more information, head over to the books section

Standalone

  • The Wretched and the Damned (2025, published by Perera Hussein)
  • The Slow Sad Suicide of Rohan Wijeratne (2024, published by Perera Hussein)
  • Numbercaste (2017, self-published and then republished by HarperCollins India)

In the Salvage Crew Universe

  • Choir of Hatred (2025, Aethon)
  • Pilgrim Machines (2024, Aethon)
  • The Salvage Crew (2020, Aethon)

    In the Commonwealth Empires trilogy

  • The Inhuman War (TBD, HarperCollins India)
  • The Inhuman Peace (2021, HarperCollins India)
  • The Inhuman Race (2018, HarperCollins India)

I also work on a fictional, open-source city called Witness over at the SciFi Economics Lab. Witness is an open-source collaborative futures design project that examines alternate economies and tries to get the underlying principles right. Then there’s Sonnet for the Subconcious War, and other poems from the Automaton (2019), a spin-off from my OSUN project - a poetry collaboration between my retrained GPT-instances and Samuel Peralta, bound for the Moon via the Lunar Codex project.

Games

I’ve contributed to Neurocracy (2021), a conspiracy thriller / murder mystery set in the pages of a rapidly mutating, near-future Wikipedia. Neurocracy won the New Media Writing Prize and was nominated for the IGF Excellence in Narrative award.

I also co-designed Witness: the Game (2021), a card game that puts you in charge of different societies facing a wave of Black Swan events (from pandemics to market crashes). You play policies depending on the type of society you run. For example, a capitalist cyberpunk world responds very differently to a religious autocracy. It draws from Witness and the work of the SciFi Economics Lab, and the Edgeryders research network in general.

Short stories

When I began writing short stories in 2017, I approached them primarily as a method to teach myself the craft. Rather than focusing merely on producing words, I wanted to examine ideas from different angles and learn to see the stories within them. Many of these works have now been collected in a print collection titled The Slow Sad Suicide of Rohan Wijeratne.

  • The Limits of my Language (2024). Published by Grimdark Magazine, issue #38. Link
  • Worrying over Potatoes (2022). Published by Google Research. Link
  • The Only Good Doctors (2021). Published by GoverningHealthFutures (a project by The Lancet x The Financial Times). Link
  • The State Machine (2020). Published on Slate.com; republished in the Big Book of Cyberpunk, edited by Jared Schurin. Link
  • Work Ethics (2020). Published on Wired.com. Link
  • Confessor (2021). Published in The Gollancz Book of South Asian Science Fiction, Volume 2, edited by Tarun K. Saint. Link
  • The Art of the Possible (2022). Published in Kalicalypse: Subcontinental Science Fiction - Fantascienza dal subcontinente. Link
  • Claws and Effect. Aboard the Lunar Codex, initially published in the Hellcats Anthology, edited by Kate Pickford. Link
  • The Tragedy of John Metcalf (2019). Published in Parallel Worlds: the Heroes Within, edited by L.J. Hachmeister. Link
  • Beatnik (2019). Published in Chronicle Worlds: Crime and Punishment. Sister story to The Writing Contest and is aboard the Lunar Codex. Link
  • A Little Bit of Kali (2019). Published in The Expanding Universe 6; republished 2021 in Escape Pod #792 (narrated by Kaushik Narasimhan). Link to republication
  • Messenger (2018). Co-written with RR Virdi. Published in The Expanding Universe 5. Nominated for the Nebula Award.
  • The Writing Contest (2018). Published in Future Visions.
  • Deep Ocean Blues (2018). Published in 2054.
  • Ozymandias (2019). Published in Horizons Beyond.
  • Omega Point: a Short Story about God (2018). Self-published. Link
  • Dreadnought (2017). Published in The Expanding Universe 3.
  • The Slow Sad Suicide of Rohan Wijeratne (2017). Self-published. Link

Nonfiction / academia

Here’s my Google Scholar profile; it contains my academic papers.

I primarily worked on natural language processing (NLP) and machine learning, with a strong focus on low-resource languages like Sinhala and Bengali. My research delves into various applications, including sentiment analysis of social media data (specifically Facebook), fake news detection and classification, and the development of language-specific resources such as corpora and sentence embeddings.

There are tendrils touching upon broader societal implications of AI, such as bias in machine learning, the challenges of moderating hate speech online, and the application of data analytics to understand socio-economic factors like poverty, regional connectivity, and even government operations. Underlying this all as a theme is the practical application of these technologies, often in the context of Sri Lanka and the challenges faced here.

In list form:

  • Wijeratne, Y., & Marikar, I. (2023). Better Question-Answering Models on a Budget. arXiv Preprint arXiv:2304.12370.
  • Weeraprameshwara, G., Jayawickrama, V., de Silva, N., & Wijeratne, Y. (2022). Sentiment analysis with deep learning models: a comparative study on a decade of Sinhala language Facebook data. Proceedings of the 2022 3rd International Conference on Artificial Intelligence in Electronics Engineering, 16–22.
  • Weeraprameshwara, G., Jayawickrama, V., De Silva, N., & Wijeratne, Y. (2022). Sinhala sentence embedding: A two-tiered structure for low-resource languages. arXiv Preprint arXiv:2210.14472.
  • Jayawickrama, V., Weeraprameshwara, G., de Silva, N., & Wijeratne, Y. (2022). Facebook for sentiment analysis: baseline models to predict Facebook reactions of Sinhala posts. International Journal on Advances in ICT for Emerging Regions (ICTer), 15(2).
  • Wijeratne, Y., Rahman, M., Hosen, K., Hossain, M., & Wahid, S. (2021). A Corpus and Machine Learning Models for Fake News Classification in Bengali.
  • Wijeratne, Y., & Attanayake, D. C. (2021). Artificial intelligence for factchecking: Observations on the state and practicality of the art.
  • Wijeratne, Y. (2021). How much bullshit do we need? benchmarking classical machine learning for fake news classification. LIRNEasia: Colombo, Sri Lanka.
  • Jayawickrama, V., Ranasinghe, A., Attanayake, D. C., & Wijeratne, Y. (2021). A corpus and machine learning models for fake news classification in Sinhala.
  • Jayawickrama, V., Weeraprameshwara, G., de Silva, N., & Wijeratne, Y. (2021). Seeking sinhala sentiment: Predicting facebook reactions of sinhala posts. arXiv Preprint arXiv:2112.00468.
  • Dias, V., Fernando, L., Amarasinghe, T., & Wijeratne, Y. (2020). Mapping Poverty and Wealth.
  • Dias, V., Lokanathan, S., & Wijeratne, Y. (2020). A Brief Primer on Bias in Machine Learning and Algorithmic Decisions.
  • Wijeratne, Y., & de Silva, N. (2020). Sinhala language corpora and stopwords from a decade of Sri Lankan Facebook. arXiv Preprint arXiv:2007.07884.
  • Wijeratne, W. (2020). Facebook, language, and the difficulty of moderating hate speech. LSE Blog, 23.
  • Wijeratne, Y., de Silva, N., & Shanmugarajah, Y. (2019). Natural language processing for government: Problems and potential. International Development Research Centre (Canada).
  • Wijeratne, Y., Samarajiva, R., Lokanathan, S., Surendra, A., & Fernando, L. (2018). Annex 15: analyzing Facebook data to understand regional connectivity.
  • Wijeratne, Y. (2018). The control of hate speech on social media: Lessons from sri lanka. CPR South.
  • Wijeratne, Y., Lokanathan, S., & Samarajiva, R. (2018). Countries of a feather: Analyzing homophily and connectivity between nations through facebook data. TPRC.
  • Samarajiva, R., & Wijeratne, Y. (2018). Use of data analytics in understanding transport.

The second port of call would be watchdog.team, where our longform explains the large systems that keep things running - combining shoe-leather journalism, data science, software engineering, and a lot of shovel work. I count these two as my most important contributions to the world.

Miscellaneous work:

  • The Ocean of Change: An exploration of megatrends within the Asia-Pacific region A mapping of inevitable changes on a 2030 timeline; regional shifts to face the challenges presented, and wildcards that can throw monkey wrenches in the works. Based on ontent analysis on national policy documents from influential governments in the region, and futures forecasts from noted bodies, and paired with statistical projections from a range of bodies such as the APERC, the UNDP, and academics in economics. Performed by LIRNEasia with funding from the UNDP.

  • Weaponising 280 characters: What 200,000 Tweets and 4,000 Bots tell us About State of Twitter in Sri Lanka 22-page report investigating the rash of bot followers haunting Sri Lankan users on Twitter after March 2018. Analyses 17 block lists, discovered some 4000 bots, and went through over 200,000 tweets. The goal was to analyze the people these bots are following – they seem to be homing in on socially active commentators on Sri Lankan politics – and to extract the logic behind how the bot network operates. We discuss theories at the end – my favourite is that there’s some company building up banks of bots across South Asia (something flagged by FT China) for potential sale later.

  • The Facebook ban in Sri Lanka: a 30,000 foot view Analyses 63,800 Facebook posts and 30,000 tweets to examine whether Sri Lanka’s March 2018 social media block actually worked. Follows with a qualitative analysis of the effects, and a note on the legalities around the matter. Last is an appendix of pages analyzed for this report.

  • Mapping Election Influence on Social Media: Part One – Twitter The 2015 general election in Sri Lanka was the first where traditional analysts started paying attention to social media. This report analyzes Twitter conversation around the 2016 election hashtags to identify nodes of influence and control in the conversation.

  • Mapping Election Influence on Social Media: Part Two – Facebook Extends the analysis onto Facebook, which is the most used social media platform in Sri Lanka, recording the growth, influence and tactics of politicians during the General Election 2015.

  • This is the Colombo Port City? An in-depth, investigative piece into the secretive Colombo Port City project, which in 2015 was still under a shroud of government-imposed secrecy. I was one of the first journalists allowed in. Contains maps, plans, theories, details, and was lauded for being one of the most informative pieces written on the Port City.