Skip to content
View canayter's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report canayter

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
canayter/README.md

Can Ayter

Bu sayfanın Türkçesi

Computational linguist and data scientist. My graduate background is in linguistics — phonetics, typology, second language acquisition — and I pair that with hands-on ML engineering to build practical language tools. The projects below aren't a separate portfolio bolted onto a resume; they're mostly direct extensions of that background, plus a couple of things I just wanted to exist.

Background

Now: Linguistic Engineer at Meta (via TEKsystems, since 2025 — Data Linguist through 2026, then Linguistic Engineer), evaluating and quality-checking Turkish-language outputs from multilingual AI models against the linguistic guidelines that shape their training data.

Before that: freelance Turkish/English translation and interpretation (legal, medical, business — ongoing since 2015); four years as a cleared linguist for U.S. federal law enforcement (KACE Company, 2019-2023) doing translation, transcription, and simultaneous interpretation; a graduate assistantship at Indiana University (2013-2015) running phonetics research and large-scale speech data collection while teaching linguistics; and, earlier, a few years coaching strength and conditioning at health clubs in Chicago and Ankara — which is exactly where Iron comes from.

Education: M.A. in Linguistics, Indiana University (Central Eurasian Studies, Hamilton Lugar School), thesis on L2 Turkish vowel acquisition — "Towards Accounting for L2 Accent: The Case of Turkish Vowel Space," an acoustic study of formant transfer in English speakers learning Turkish, advised by Dr. Öner Özçelik. An Erasmus exchange year at Uppsala University in Turkic linguistics and philology. B.A. in Linguistics from Hacettepe University. A Data Science & Visualization certificate from Northwestern, more recently.

The full resume, with dates and specifics, is at ayter.com.

The Network

Small, focused tools, most of them tracing back to something above rather than a business plan. What's actually built and live versus described-but-not-yet is called out plainly in each project's own README — not everything here is equally finished, since these are mostly built solo, alongside everything else.

Language & linguistics — where the linguistics background shows up most directly:

  • Langram (live) — a Turkish course for English speakers, built on a real morphophonological engine: exercises are generated live from grammar rules, not drawn from a fixed list.
  • Rosetta (live) — an interactive map of 500+ world languages, typology, endangerment, and mutual intelligibility, sourced from WALS, Ethnologue, Glottolog, and PHOIBLE.
  • Semaphore (live) — real-time ASL fingerspelling recognition in the browser, plus a Claude-assisted text-to-sign mode.
  • Cipher (live) — translation across five languages with real part-of-speech tagging and a purpose-built Turkish morphological analyzer.
  • Timbre (live) — the direct continuation of the thesis above: a vowel-space research writeup and reference chart, with live formant extraction as the planned next step.

Body & mind — where the coaching background shows up:

  • Iron (live) — a training and nutrition platform, every calculator cited back to the study behind it.
  • Reverie (live) — binaural beats, breathing protocols, and guided meditation across six traditions, explained on their own terms.
  • Nous (live) — cognitive training on named tasks (Raven's Matrices, Stroop, N-back), an ADHD hub, and a wellness screening with real crisis resources.
  • Forage (live) — an AI pantry companion that shares nutrition data live with Iron.

Money:

  • Prudence (live) — a cited, client-side guide to debt payoff order, debt vs. invest, and refinance math. (Repo is private.)

Just for fun:

  • Vault (live) — eight games in a neo-retro arcade shell, seven real classics plus one joke original.
  • Cadence (live) — turns Last.fm history into a threaded Bluesky post, with an optional self-hosted weekly auto-scheduler.

Can Ayterayter.com · LinkedIn

Pinned Loading

  1. MADAIN MADAIN Public

    Forked from AEKoller/MADAIN

    Using TensorFlow and Keras to predict skin cancer from mole images, achieving high accuracy and precision through extensive testing and benchmarking of CNN architectures and optimization techniques.

    Jupyter Notebook

  2. iron iron Public

    Full-stack training platform — anatomy-linked exercise atlas, workout builder, nutrition planner, and evidence-based calculators. Single-file, zero backend.

    HTML 1

  3. timbre timbre Public

    Vowel-space and formant analysis, grounded in a Master's thesis on L2 Turkish vowel acquisition. Cardinal vowel chart + 41-language reference data.

    Python 1

  4. rosetta rosetta Public

    Interactive map of 500+ world languages — typology, family trees, endangerment status, mutual intelligibility, all sourced from WALS/Ethnologue/Glottolog/PHOIBLE.

    JavaScript

  5. credit-risk-classification-analysis credit-risk-classification-analysis Public

    A machine learning model using logistic regression to predict loan defaults based on financial data. The purpose of the analysis was to assist in risk assessment and decision-making processes in th…

    Jupyter Notebook 1

  6. wine_and_climate_change wine_and_climate_change Public

    Forked from AEKoller/wine_and_climate_change

    Utilized Pandas for data cleaning and integration to create a high-quality, uniform dataset of global wine production and historical temperature records. Employed SciPy for statistical analysis to …

    Jupyter Notebook