Meghdad Farahmand, Developer in Berlin, Germany
Meghdad is available for hire
Hire Meghdad

Meghdad Farahmand

Verified Expert  in Engineering

Artificial Intelligence (AI) Developer

Location
Berlin, Germany
Toptal Member Since
June 6, 2019

Meghdad has more than 15 years of experience in AI, including developing cutting-edge NLP models for international clients such as Siemens, Unilever, and Allianz. He has a strong theoretical background that includes a bachelor's in computer engineering, a master's in computer science with a specialization in AI, and a Ph.D. in computer science with a specialization in NLP.

Portfolio

Atreus GmbH
Python, Data Lakes, Natural Language Processing (NLP)...
Schultz Family Foundation
Natural Language Processing (NLP), Large Language Models (LLMs), Python...
Archipelo Inc.
Machine Learning, Natural Language Processing (NLP), Data Lakes, ETL...

Experience

Availability

Full-time

Preferred Environment

PyCharm, IntelliJ IDEA, Visual Studio Code (VS Code), Linux, MacOS

The most amazing...

...accomplishment was being an architect and lead developer of an ML pipeline of a software support tool that allowed ML model deployment at scale.

Work Experience

Senior Data Science Consultant

2022 - PRESENT
Atreus GmbH
  • Created ML/NLP features to support business consulting use cases.
  • Developed ML/NLP features to support business marketing and sales use cases.
  • Developed LLM-based POCs for the future of the business.
Technologies: Python, Data Lakes, Natural Language Processing (NLP), Artificial Intelligence (AI), Strategy, Data Science, Large Language Models (LLMs), Amazon Web Services (AWS), Open-source LLMs, Minimum Viable Product (MVP), Generative Artificial Intelligence (GenAI), Language Models

Senior Data Science Consultant

2021 - 2022
Schultz Family Foundation
  • Developed a conversational AI (before ChatGPT and in 2021) that answers entrepreneurial questions by consulting high-quality materials such as well-known books and articles.
  • Developed various ML and NLP applications such as text classifiers, personality classifiers, time-series predictors, and business failure analysis systems.
  • Developed an analytical and predictive dashboard on top of emails, social media, reviews, and notes to help entrepreneurs with their business journey.
Technologies: Natural Language Processing (NLP), Large Language Models (LLMs), Python, Google Cloud, Data Science, Data Analysis, Solution Architecture, Open-source LLMs, Minimum Viable Product (MVP), Generative Artificial Intelligence (GenAI), Language Models

ML Lead

2021 - 2022
Archipelo Inc.
  • Led the ML and NLP development efforts in building an intelligent code discovery platform for software developers.
  • Guided the development of an ETL pipeline that deals with large datasets.
  • Participated and enjoyed collaborating on the most detailed development components, as a hands-on and participating team lead.
Technologies: Machine Learning, Natural Language Processing (NLP), Data Lakes, ETL, Software Architecture, Software Engineering, Test-driven Development (TDD), Python, Data Science, Named-entity Recognition (NER), Solution Architecture, Open-source LLMs, Language Models

NLP Lead — Senior ML Engineer

2021 - 2021
Merantix
  • Consulted clients as well as internal stakeholders about NLP projects.
  • Developed some components and modules within the MLOps infrastructure.
  • Created a roadmap for developing NLP software components, packages, and pipelines concerning market signals.
  • Developed a label analysis package to evaluate the quality of the labels in labeled data.
Technologies: Python 3, Google Cloud Platform (GCP), PyTorch, Docker, Software Architecture, Machine Learning Operations (MLOps), OOP Designs, Data Science, Solution Architecture, Language Models

Senior Data Scientist

2018 - 2020
omni:us
  • Developed and deployed a highly scalable and customizable document classification pipeline for multilabel and multiclass classification based on deep neural networks.
  • Studied and integrated transfer learning anomaly detection and multitask learning in order to improve the performance of the deployed classification and existing named entity recognition (NER) micro-services.
  • Carried out statistical, distributional, and syntactic analysis of textual data to improve downstream models' performance.
  • Co-advised M.Sc theses on NER and active learning.
  • Developed a text-to-text mapping service based on a siamese neural network.
Technologies: Docker, Deep Learning, TensorFlow, PyTorch, Python, Data Science, Statistical Modeling, Named-entity Recognition (NER), Language Models

Data Scientist

2017 - 2018
Market Logic Software
  • Developed and deployed deep learning-based NLU and NLG services for a chatbot and a reinforcement learning-based dialogue manager for some consumer goods and technology world leaders.
  • Developed and deployed a text classification pipeline with active learning.
  • Developed a data analytics prototype for extracting trends and hidden marketing insights from market research documents.
  • Researched question answering, NLG, and abstractive summarization and studied the feasibility and performance of cutting-edge models in production.
  • Researched and studied the feasibility of different methods for leveraging huge incoming client inputs to train the existing models in real time.
Technologies: Reinforcement Learning, Deep Learning, PyTorch, Python, Data Science, Named-entity Recognition (NER), Amazon Web Services (AWS), NoSQL

NLP Engineer

2011 - 2012
UNDL Foundation
  • Researched and developed statistical models for extraction and alignment of idiomatic expressions across nine European languages.
  • Employed high-performance computing to handle big data.
  • Carried out research and published scientific articles.
Technologies: MATLAB, Java, Data Science, Statistical Modeling, Data Analysis, Language Models

Wordview

https://github.com/meghdadFar/wordview
Wordview is a Python package for exploratory data analysis of text. It provides many statistics about your data in the form of plots, tables, and descriptions, giving you a high-level and detailed overview of your data. It has functions to analyze explicit text elements such as words, n-grams, POS tags, and multi-word expressions, as well as implicit elements such as clusters, anomalies, and biases.

Bourbon

https://github.com/meghdadFar/bourbon
Bourbon is a Python package for general reinforcement learning (RL). It can be used to train an RL agent for any problem that can be mapped to RL. If the issue at hand satisfies the main RL requirements, then it can be solved via RL, and hence, Bourbon can be the right solution.

Languages

Python, Java, C++, Python 3, SQL

Tools

Named-entity Recognition (NER), GitHub, Shell, MATLAB, MongoDB Shell, AWS CLI, Seaborn, PyCharm, IntelliJ IDEA, Jenkins, ABBYY

Paradigms

Data Science, ETL, Test-driven Development (TDD)

Other

Deep Learning, Multivariate Statistical Modeling, Distributional Semantics, Natural Language Understanding (NLU), Natural Language Processing (NLP), Semantic Composition, Word2Vec, Machine Learning, Artificial Intelligence (AI), Statistics, Statistical Modeling, Data Visualization, Data Analysis, Analytics, GPT, Generative Pre-trained Transformers (GPT), Large Language Models (LLMs), Generative Artificial Intelligence (GenAI), Language Models, Reinforcement Learning, Regression, Clustering, Mixture Models, Software Architecture, OOP Designs, Cloud Computing, Software Engineering, Predictive Analytics, Planning, Clustering Algorithms, Solution Architecture, Open-source LLMs, Minimum Viable Product (MVP), OCR, Machine Learning Operations (MLOps), Cryptocurrency APIs, Project Leadership, Leadership, Strategy

Libraries/APIs

PyTorch, Natural Language Toolkit (NLTK), NumPy, Scikit-learn, Pandas, SpaCy, TensorFlow

Platforms

Amazon Web Services (AWS), Linux, MacOS, Amazon EC2, Docker, Google Cloud Platform (GCP), OS X, Visual Studio Code (VS Code), Azure

Storage

MongoDB, Amazon S3 (AWS S3), Data Lakes, Google Cloud, NoSQL

2012 - 2017

PhD Degree in Computer Science

University of Geneva - Geneva, Switerland

2008 - 2010

Master's Degree in Computer Science

University of Lugano - Lugano, Switzerland

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring