Index

This page catalogs every page in Sadra’s wiki version of his CV. Browse the sections in the sidebar, or follow the graph to see how pages connect.

Profile

• Open Source Philosophy — Open-sourcing NLP research is what got the field to things like ChatGPT. The same openness now makes it very…
• Research Agenda — The work is about how AI fits into the way people already work, rather than about the models themselves. One…

Education

• Bachelor of Electrical Engineering and Computer Science — Sharif University of Technology (2023) — The bachelor’s thesis became Docalog: Multi-Document Dialogue System Using Transformer-Based Span Retrieval, a…
• Master’s of Computer Science — University of Southern California (2025) — Earned during the PhD program.
• PhD in Computer Science — University of Southern California (2028) — Research at the intersection of HCI and NLP, advised within the Adaptive Computing Experience (ACE) Lab with a…

Research Experience

• Sharif University of Technology — Speech and Language Processing Lab (2021) — Two lines of work, both pre-ChatGPT
• Thread: AI Aid for Decision Making (2024) — How LLM presentation style influences the decisions people make, across two studies
• Thread: AI for Education (2024) — How AI systems explain things to people with different levels of expertise, across three efforts
• Thread: Human-AI Interaction in Knowledge Work (2023) — How people who already have a job to do trust, oversee, and integrate AI inside the tools they work in.…
• University of Sydney, Research Assistant Intern — Immersive and visual computing: augmented reality, virtual reality, and holography, worked on alongside…
• USC ISI — CUTE LAB NAME (2025) — Secondary lab affiliation alongside the ACE Lab, bringing the NLP side of the work: the two labs together host…
• USC — Adaptive Computing Experience (ACE) Lab (2023) — Primary research home for the PhD — the Adaptive Computing Experience (ACE) Lab, Souti Chattopadhyay’s lab at…

Industry Experience

• Asr Gooyesh Pardaz (2020) — Speech and conversational AI for commercial products
• Microsoft PROSE — Applied Scientist Intern (2026) — Sadra spent the summer of 2026 as an applied scientist intern on the Microsoft PROSE (PROgram Synthesis using…
• Open Science Laboratory (OpenSciLab) (2019) — Co-founded with a group of friends, and led development of open-source scientific tooling. The motive was plain…

Publications

• Auditing and Controlling AI Agent Actions in Spreadsheets (2026 · under submission) — How users audit and retain control over AI agents acting inside spreadsheets. First-author work; co-authored…
• Cognitive Biases in LLM-Assisted Software Development (ICSE · 2026) — An observational study built on watching 36 developers actually work. It found that 48.8% of their programming…
• Did You Think This Through?: Effect of Varying Advice Styles on Users’ Performance and Contemplation (VL/HCC · 2026) — A survey of 98 participants verifying factual claims with LLM advice presented in different rhetorical styles.…
• How LLMs’ Anthropomorphism and Sycophancy Compromise Expert Evaluation of Therapeutic AI (2026 · under submission) — Examines how anthropomorphic and sycophantic model behavior distorts expert judgment of therapeutic AI systems…
• Prompting Reflection: Encouraging Users to Critically Evaluate AI Suggestions (2026 · under submission) — Designs reflection-prompting strategies that push users toward more deliberate decisions rather than accepting…
• O (JOSS · 2026) — The paper treats saving a machine learning model as a safety and transparency problem rather than a convenience…
• Synthesizing Program Analyzers to Help Programmers Answer Questions About Code (VL/HCC · 2026) — A system bridging LLMs with CodeQL so novice programmers can interrogate large codebases. Uses RAG-based…
• Beyond the Page: Enriching Academic Paper Reading with Social Media Discussions (UIST · 2025) — A paper is formal and finished, while the actual conversation about it happens on social media. This work…
• ELI-Why: Evaluating the Pedagogical Utility of Language Model Explanations (ACL · 2025) — A good explanation is not one thing. Explaining the same thing to a fifth grader and to a graduate student are…
• Exploring the Challenges and Opportunities of AI-assisted Codebase Generation (VL/HCC · 2025) — A study of “vibe-coding”: building software by prompting an LLM over and over instead of writing the code…
• PahGen: Generating Ancient Pahlavi Text via Grammar-Guided Zero-Shot Translation (LoResMT@NAACL · 2025) — Pahlavi (Middle Persian) barely exists in digital form, and a language with no data is a language on its way…
• ParsiPy: NLP Toolkit for Historical Persian Texts in Python (ALP@NAACL · 2025) — Historical languages are hard to work with, and for several reasons at once: the orthography is complicated,…
• Persian in a Court: Benchmarking VLMs in Persian Multi-Modal Tasks (EvalMMG · 2025) — A benchmark evaluating vision-language models on Persian multi-modal tasks.
• ReaxFF Parameter Set for Boron Clusters and Icosahedral Boron Crystals (Journal of Physical Chemistry C · 2025) — Icosahedral boron is a candidate material for semiconductors and energy storage, but the synthesis conditions…
• Samila: A Generative Art Generator (arXiv · 2025) — Samila makes images by scattering many thousands of points, each one placed by a formula with random…
• Trust Dynamics in AI-Assisted Development: Definitions, Factors, and Implications (ICSE · 2025) — The paper starts from a question that sounds simple and isn’t: what do developers actually mean when they say…
• A Semidefinite Relaxation Approach for Fair Graph Clustering (KDD (Ethical AI Workshop) · 2024) — A semidefinite relaxation for clustering graphs under fairness constraints — addressing discriminatory social…
• Nafas: Breathing Gymnastics Application (arXiv · 2024) — Technical report on Nafas, a set of breathing exercises for people at a computer for long hours, paired with…
• Representative Sample Size for Estimating Saturated Hydraulic Conductivity via Machine Learning (Water Resources Research · 2024) — Hydrology adopted machine learning widely but rarely examined how data heterogeneity and sample size affect…
• SynTran-fa: Generating Comprehensive Answers for Farsi QA Pairs via Syntactic Transformation (Preprints.org · 2024) — Data augmentation for Farsi question answering: converts short QA pairs into full, complete answers by…
• naab: A Ready-to-Use Plug-and-Play Corpus for Farsi (JAIAI · 2024) — Lower-resource languages like Farsi need large training data, and at the time the work was done there was even…
• A Review of the Recent Speech Recognition Methods (JVS · 2023) — A literature review of contemporary speech recognition methods, growing out of the Farsi ASR and TTS systems…
• A Literature Review on Rater Agreement Metrics (pycm.io · 2022) — Survey of rater agreement metrics, written to give PyCM users a clearer basis for choosing among them — part of…
• Docalog: Multi-Document Dialogue System Using Transformer-Based Span Retrieval (DialDoc@ACL · 2022) — The team’s entry for the DialDoc-22 (MultiDoc2Dial) shared task, and the substance of Sadra’s BSc thesis.…
• Experimental Dataset of Electrochemical Efficiency of a Direct Borohydride Fuel Cell (DBFC) (ChemRxiv · 2022) — A measurement release rather than a modelling paper: the record of impedance and polarization tests run on a…

Projects

• Art — ASCII Art Library for Python (2019) — Art does the “smart” placement of typed special characters and letters so that together they form a visual…
• Drux — Drug Release Analysis Framework (2025) — Drux simulates and plots drug release profiles from mathematical models — a simple, extensible, reproducible…
• IPSpot — System IP Address Fetcher (2025) — IPSpot tells you your system’s IP address and where it looks like you are. It reports the private address the…
• Memor — Conversational Memory Across LLMs (2025) — Manages the memory of a user’s interactions with LLMs. Users can tap the history of past conversations when…
• MyTimer — A Timer for Command Line Enthusiasts (2021) — MyTimer is a simple but comprehensive timer for terminal users: you set it directly from the command line, so…
• Mycoffee — Coffee Brewing Recipes for Terminal Users — A community-based coffee recipe CLI application. Well received by the programming community on social networks,…
• Nafas — Breathing Gymnastics Application (2021) — Nafas is a collection of breathing exercises for people who have been sitting at a computer for too many hours.…
• Nava — OS-Native Sound Engine in Python (2023) — Nava plays sound in Python with no dependencies and no platform restrictions, on Windows, macOS, and Linux.…
• OPEM — Open Source PEM Fuel Cell Simulation Tool (2019) — OPEM, the Open-Source PEMFC Simulation Tool, evaluates how a proton exchange membrane fuel cell performs and…
• OPR — Optimized Primer Design Tool (2024) — OPR designs, validates, and optimizes primers for PCR, qPCR, and sequencing, consolidating primer design…
• PyCM — Multi-Class Confusion Matrix Library (2019) — PyCM evaluates classifiers after the fact: it works from the predictions a model has already made and supports…
• O (2024) — A flexible I/O interface for ML pipelines, built around one problem: saving a model with pickle means shipping…
• PyRGG — Python Random Graph Generator (2019) — PyRGG makes up random graphs for network simulation: it synthesizes graph data for people who need a network to…
• Samila — Generative Art Generator (2021) — Samila is a Python prototyping tool for artists: it makes an image by scattering many thousands of points, each…
• Sharif-Wav2Vec2.0 — Farsi Speech Recognition Model (2022) — A Wav2Vec2.0 speech-processing model tailored for Farsi: the base model, fine-tuned on 108 hours of Farsi audio…
• ToCount — Lightweight Token Estimator (2025) — Estimates token counts for LLM input using rule-based and ML methods. Built for prompt analysis, token…
• XNum — Universal Numeral System Converter (2025) — Converts digits across numeral systems — English, Persian, Hindi, Arabic-Indic, Bengali, and others.…
• naab — Farsi Text Corpus (2022) — A large, cleaned, ready-to-use open-source Farsi text corpus: ~130GB, 250 million paragraphs, 15 billion words.…

Honors and Awards

• NLnet Foundation Grant — PyCM (2025) — Funding to develop distance/similarity metrics and benchmarking in PyCM — statistical analysis of machine…
• PSF Software Development Grant — Art (2024) — Funding to add multi-line and custom font features to the Art library.
• PSF Software Development Grant — Nava (2025) — Funding to develop OS-based sound engines for the Nava library. Part of the PSF’s support for Python packages…
• Tinker Research Grant — Thinking Machines Lab (2026) — For fine-tuning and post-training large language models. The funded project is on Persian prosody-informed LLMs.
• Trelis AI Micro-Grant — PyCM (2024) — A development micro-grant to build an accessible RESTful API for PyCM, making the library usable by people not…
• USC Presenter for ViterbiTrek — Selected as a USC presenter at ViterbiTrek, a Viterbi program connecting students with industry professionals.…
• Vector Scholarship in AI — Recognizes exceptional candidates in Vector Institute-recognized or AI-focused study paths.

Services

• Peer Review —

Teaching and Outreach

• NLP and Speech Processing Workshop — Sharif University of Technology — Taught a session on natural language processing and speech processing.
• Section Leader — Stanford Python Program — Section leader for a global program teaching the Python programming language.

Talks

• PyCon US 2026 — Memor Poster (PyCon US · 2026) — Presented Memor — managing and transferring conversational memory across LLMs.

Skills

• Professional Skills — Collaboration and leadership: Teamwork, leadership, project management
• Research Methods — Study design: Research design, user interviews, surveys and questionnaire design, A/B testing,…
• Technical Skills — Programming languages: Python, C, C++, Java, JavaScript, TypeScript, HTML, CSS, MATLAB, Assembly,…

References

• Jonathan K. Kummerfeld — Affiliation: Assistant Professor of Computer Science, University of Sydney Relationship: Co-author on the…
• Jonathan May — Co-author on the advice-styles and reflection-prompting work, and on therapeutic AI.
• Nenad Medvidovic — Affiliation: Professor of Computer Science, University of Southern California Relationship: Co-author on the…
• Sepand Haghighi — Co-author or co-maintainer on PyCM, Art, Samila, Nafas, PyRGG, PyMilo, MyTimer, and the therapeutic AI study.
• Souti Chattopadhyay — Co-author on the majority of the HCI work — trust dynamics, cognitive biases, codebase generation, reflection…
• Sumit Gulwani — Affiliation: Microsoft PROSE team Relationship: Hosted and worked with Sadra during his summer 2026 applied…