Language · People · AI quality

I turn complex information into clear guidance, and check that it's right.

Social worker and psychology graduate with a linguistics background. Ten-plus years of assessing, training and supporting people, and since 2024, using AI tools for research, fact-checking, text analysis and learning design.

  • Hebrew: native
  • English: C2
  • Spanish: advanced
10+ yearsassessing, training and supporting people
80+international client processes managed end to end
20–40support agents guided and evaluated per shift

Selected work

Three questions run through these projects: Where did this come from? What needs fixing? How do I explain it clearly?

AI-assisted analysis of a multi-year text corpus

2026
  • Research
  • AI output review
  • Qualitative coding
Challenge
A large, multi-year body of personal text, too big to read by hand. LLMs can summarize it, but their output isn't reliable unless it's checked.
What I did
Designed a multi-phase analysis with Claude Code, Claude and ChatGPT. Built an inductive taxonomy and a systematic coding framework, wrote the instructions for each tool, checked their output against the source text, and ran robustness tests on the findings.
Outcome
Validated datasets, visualizations and written reports, plus a repeatable method for using LLMs on qualitative data without trusting them blindly.

The source data is personal, so this project is shown as a method only.

Ethical dilemma simulator

2026
  • Scenario-based learning
  • Hebrew UX
  • Built with Lovable
Challenge
Preparing for an open-book ethics exam in psychology means applying principles to messy cases, not memorizing them.
What I did
Designed a scenario bank and a four-stage practice flow: identify the type of dilemma and the conflict, choose the relevant sources, consult a simulated colleague, and write a reasoned decision that receives feedback. I owned the content structure, the scenarios, the interaction sequence and the Hebrew feedback, and added a clear disclaimer that it is a learning tool, not professional advice.
Outcome
A working practice tool, in Hebrew, that walks the learner from spotting the dilemma to defending a decision.

Course material turned into interactive learning

2026
  • Learning design
  • Assessment
  • Content structuring
Challenge
Dense course material (a full CBT course, or particle physics) is hard to review and practice from lecture notes alone.
What I did
Converted a CBT course into a self-paced Hebrew learning tool with nine units, a glossary, clinical simulations and a 30-question practice exam with explanations after each answer. Applied the same approach to an interactive particle-physics explainer, breaking complex material into short, navigable units.
Outcome
Self-paced tools a learner can explore in their own order and use to test themselves.

Kabardian Morphology Lab

2026
  • Linguistics
  • Interactive learning
  • Hebrew
Challenge
Explain to Hebrew speakers how a polysynthetic Caucasian language packs a whole sentence (agent, recipient, object, tense) into a single verb.
What I did
Built an interactive Hebrew-language page around documented example sentences: clickable morphemes, a “build the verb” exercise that adds one grammatical layer at a time, and competing scholarly analyses (ERG vs. OBL) shown side by side instead of hidden.
Outcome
A hands-on explainer that turns an abstract topic in morphology into something a learner can see and try.

Token Field Guide

2026
  • AI tools research
  • Source evaluation
Challenge
Advice on using AI models efficiently is scattered across vendor docs, papers and forums, and much of it is folklore.
What I did
Researched and synthesized official documentation, papers and community sources into a practical guide: which model and effort level to use for which job, where tokens actually go, and how writing in Hebrew affects cost. Every tip is labeled by evidence level: official, community, unverified or debunked.
Outcome
Eight ranked rules, per-platform notes and a copy-paste kit, with each tip's evidence level visible.

Onboarding, evaluation and quality at scale

2019–2023
  • Training
  • Quality assurance
  • Feedback
Challenge
Bring new people up to standard fast, in a high-volume tech operation (Wolt) and in a teaching organization (New School).
What I did
At Wolt, guided 20–40 agents per shift against CSAT and TPH targets, reviewed conversations to find knowledge gaps, onboarded new hires and wrote structured evaluations that informed continued-employment decisions. At New School, oversaw teaching quality for 16 instructors, built training materials for recurring issues and gave one-to-one feedback.
Outcome
Promoted from customer support to Performance Lead and Onboarding Trainer within six months.

Internal company material isn't shared here.

What I bring

The common thread: understanding people and language well enough to judge whether something is clear, accurate and useful.

AI output review and data work

Fact-checking against primary sources, text labeling and classification, guideline adherence, writing instructions for AI tools.

Research and analysis

Research methods and statistics, qualitative coding, taxonomy building, multi-source synthesis, analytical writing.

Language

BA in Linguistics (semantics, syntax, phonology, sociolinguistics). Professional writing, editing and proofreading in Hebrew and English.

Training and learning design

Onboarding, workshop facilitation, learning-needs assessment, training materials, practice exams, feedback and coaching.

Assessment

Structured interviews, psychosocial and vocational assessment, written evaluations and professional reports.

Tools

Claude, Claude Code, ChatGPT, Lovable, Salesforce, Intercom, Slack, Google Workspace, Microsoft Office.

Background

The full story is in my CV.

  • 2024 – nowIndependent projects in language, content, research and AI
  • 2024 – 2025Client Manager · Tammuz Family Surrogacy
  • 2022 – 2023Performance Lead / Onboarding Trainer, previously Customer Support · Wolt
  • 2020 – 2022Team Lead and Social Worker · Ramat Gan Municipality
  • 2017 – 2019Employment Social Worker · Tel Aviv-Yafo Municipality; QA Manager, Teaching Quality · New School
  • EducationBA Psychology · Tel Aviv University (93); BA Social Work and Linguistics · Hebrew University (91); MA studies in Rehabilitation Psychology · Ariel University