Skip to content
View maxschmaltz's full-sized avatar

Highlights

  • Pro

Block or report maxschmaltz

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.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

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

Report abuse
maxschmaltz/README.md

I'm Max, nice to have you here👋


About Me📌

  • Young and enthusiastic NLP / LLM Engineer
  • 5+ years of experience in Python, Machine and Deep Learning, Natural Language Processing (including Large Language Models)
  • Generative LLMs? Bring it on!
  • Creating tools for German NLP as a hobby
  • Learn faster than Logistic Regression
  • Take a look at my resume (might render incorrectly in Safari)

Connect🤗

image image Calendly image image


Tech Stack🛠️

Programming Languages

image image image Java

Python: AI Basic Frameworks

image image image

Python: NLP

🦜️🔗LangChain OpenAI API 🦙llama-cpp-python 🤗 Transformers 🤗 Datasets NLTK spaCy Gensim pynini

Python: TSF

statsmodels pmdarima XGBoost

Python: Misc

image image image Matplotlib image image

Deployment

image Docker Hub image


Professional Background🧑‍💻

IAV: Working Student for LLMs

    Further deepening and extension of my competencies | 2023-present | Berlin, Germany | Partially remote

    • Implemented and integrated a tool for evaluation of Large Language Models and Retrieval-Augmented Generation pipelines (Python, LangChain, Ragas, HF datasets, Azure)

    • Currently, connecting OpenAI function calling to our internal on-premise models (Python, TGI, LangChain)

MKSKOM: Data Scientist / NLP Engineer

    Best boost for my competencies | 2021-2023 | Moscow, Russia | Remote

    • Implemented backend for a custom Llama-2-based Retrieval-Augmented Generation Engine for a large customer (Python, LangChain, Llama.cpp, HF Transformers)

    • Increased the number of responses from potential customers on a freelance platform by 5 times by independently designing and implementing an automated LLM-based tool for search and filtering posts and contacting potential customers with relevant infos (Python, Puzzle [internal tool])

    • Implemented tools for Natural Language Processing, Time Series Prediction, Optimization Modelling, Data Analysis (Python, PyTorch, spaCy, scikit-learn, statsmodels, Pandas, NumPy)

    • Сommunicated with potential customers

Eberhard Karls Universität Tübingen: Student Assistant, Tutor (Various Courses)

    Best boost for my communication | 2022-2023 | Tübingen, Baden-Württemberg, Germany | On-site

    • Created Finite-States Transducers for measuring distance between German dialects for the course "String Algorithms" (Python, pynini)

    • Contributed into holding lectures, created and evaluated assignments for courses "Python for Beginners", "Statistical Language Processing II", "String Algorithms"

Yandex: Assessor

    Deeper understanding of various IT topics | 2021-2022 | Moscow, Russia | Remote

    • Evaluated search engine results on IT-themed queries

    • Evaluated machine translations

Lomonosov MSU Gymnasium: Course Instructor (Linguistics for Olympiades)

    First experience of lecturing | 2021-2022 | Moscow, Russia | On-site

    • Held lectures, composed and evaluated assignments for the Course "Linguistics for Olympiads"


Educational Background🧑‍🎓

    Computational Linguistics | 2022-2024 | Tübingen, Baden-Württemberg, Germany

    • 2024 (in progress) Bachelor thesis on applying LLMs to solving non-trivial linguistic tasks (on the example of splitting German compounds (Python, LangChain, PyTorch, HF transformers, spaCy)

    • 2024 Group project on investigating influence of RL fine-tuning data on biasing LLMs (Python, HF transformers)

    • 2023 Participated at the SemEval 2023 and published a paper at the ACL Anthology (Python, HF transformers, HF datasets)

    Computational Linguistics | 2022 | Tübingen, Baden-Württemberg, Germany

    • 2022 Applied for the Bachelor at the Uni Tübingen, got admitted and transferred

Lomonosov MSU: Bachelor (incomplete)

    Theoretical and Applied Linguistics | 2019-2022 | Moscow, Russia

    • 2022 Personal project DERBI: a tool for automatic inflection of German words (Python)

    • 2022 Conference diploma for DERBI: Lomonosov-2022 (Lomonosov Moscow State University), Science Sessions-2022 (Kant Baltic Federal University), XXII International Conference of Young Slavists (Tallinn University)

    • 2021 Term paper on predicting ablaut class of German strong verbs (Python, PyTorch)

    History and Philology | 2017-2019 | Moscow, Russia

    • 2019 Prize-winner of All-Russian Olympiad for Linguistics (gives no-exam admission to the top university of choice)

    • 2019 Graduation with a Gold Medal


Fun Facts😎

  • I've been living on my own from the age of 15
  • I gave up Lomonosov MSU to leave for Germany in 2022
  • My best friend and I have found a startup, the app is currently in beta testing
  • In 2019, I became a prize-winner of All-Russian Olympiade for Linguistics, while I had only 5 month to prepare
  • The prize gave me a right to enter any university of my choice in Russia for major Linguistics without any exams at all
  • I completed a musical education, now I compose songs from time to time
  • I got my first driving license when I was 17: that one was for motorcycles with 125cc and smaller engines
  • At 18, came back to the same driving school to obtain further licenses: for autos and for large displacement motorcycles; the funny part is, I drove to the exam place on my moto to take an exam for driving motos
  • I learned to weld in 3 days just for fun and crafted a food stand for my dog

Popular repositories Loading

  1. DERBI DERBI Public

    DERBI (DEutscher RegelBasierter Inflektor) is a simple rule-based automatic inflection model for German based on spaCy. Applicable regardless of POS!

    Python 3 2

  2. MarChie MarChie Public

    An Open Source Tool for Analyzing Discrete Markov Chains.

    Jupyter Notebook 1

  3. WebSemble WebSemble Public

    An ensemble approach to solution of Clickbait Challenge at SemEval 2023.

    Python

  4. DummyLyricsGenerator DummyLyricsGenerator Public

    A dummy web app for generating funny lyrics with ChatGPT.

    Python

  5. Hirer Hirer Public

    A Simple LLM-Powered Hiring Plan Creator.

    Python

  6. maxschmaltz maxschmaltz Public

    I'm Max, nice to have you here👋