GenAI Consultant
QPharma, USA
Keywords
Skills
Azure Container Apps/Instances, Machine Learning, Data Factory, Function Apps, Blob | AWS Lambda, S3, RedShift,
SageMaker | Python | PySpark | Dask | Ray | SQL | Airflow | Google Analytics, BigQuery, Data Studio, Bucket | Git
CI/CD Power BI | Django, Flask, FastAPI | Angular Typescript | PostgreSQL | Pytorch | Hugging Face | LangChain
NLTK Gensim | SpaCy | OpenAI | Jira Kanban Confluence | Bagging | Boosting| Clustering | Deep Neural Networks
Explainable AI (SHAPLEY, SVM, etc.) | Bayesian Optimization | Time Series Models (GARCH, ARIMA, etc.), GlounTS
Transformers (Hugging Face) | RAG | Langchain | vLLM | LlamaIndex | Agents | QLoRA | Vector Database Pine
SageMaker | Python | PySpark | Dask | Ray | SQL | Airflow | Google Analytics, BigQuery, Data Studio, Bucket | Git
CI/CD Power BI | Django, Flask, FastAPI | Angular Typescript | PostgreSQL | Pytorch | Hugging Face | LangChain
NLTK Gensim | SpaCy | OpenAI | Jira Kanban Confluence | Bagging | Boosting| Clustering | Deep Neural Networks
Explainable AI (SHAPLEY, SVM, etc.) | Bayesian Optimization | Time Series Models (GARCH, ARIMA, etc.), GlounTS
Transformers (Hugging Face) | RAG | Langchain | vLLM | LlamaIndex | Agents | QLoRA | Vector Database Pine
Project history
03/2024
-
Present
Main Responsibilities
▪ Implemented a RAG approach together with an LLM to perform content management.
▪ Implemented a RAG based chat model to help resolve customer queries of an online gambling platform.
▪ Implemented a RAG approach together with an LLM to perform content management.
▪ Implemented a RAG based chat model to help resolve customer queries of an online gambling platform.
02/2024
-
03/2024
R&D Consultant on S2ST, TTS, ASR
Qonda GmbH, DE
Wrote the two research proposals to be presented to Investment Bank Berlin (IBB):
I. Flow normalization using multi-stage single-channel speech enhancement approach in ASR.
II. Optimizing S2ST: Model Compression and Data Augmentation for Efficient Speech-to-Unit Translation.
I. Flow normalization using multi-stage single-channel speech enhancement approach in ASR.
II. Optimizing S2ST: Model Compression and Data Augmentation for Efficient Speech-to-Unit Translation.
10/2023
-
02/2024
GenAI Consultant
NDI GmbH, CH
(Internet and Information Technology, < 10 employees)
Main Responsibilities
▪ Implemented a RAG approach together with an LLM to perform content management.
▪ Implemented a RAG based chat model to help resolve customer queries of an online gambling platform.
▪ Implemented a RAG approach together with an LLM to perform content management.
▪ Implemented a RAG based chat model to help resolve customer queries of an online gambling platform.
04/2023
-
07/2023
SPECIALIZED CHAT BOT FOR A INSURRANCE COMPANY
Anonym
(Insurance, 500-1000 employees)
Fine tuning of an LLM (meta/Llama_2) to improve customer queries and Q&A. Respond more consistently, learning focus, reduce hallucination, gain knowledge on insurance data. Vector averaging of word embeddings was used to measure model performance.
Llama_2 | Hugging Face | Pytorch | Embedding Distance | LAMINI | FastAPI | Pandas | Numpy
Llama_2 | Hugging Face | Pytorch | Embedding Distance | LAMINI | FastAPI | Pandas | Numpy
01/2023
-
03/2023
MARKET RESEARCH ON LLMS
Ministry of higher education Pakistani
(Internet and Information Technology, 1000-5000 employees)
Summary review on LLM concepts, including private and open-source models, comparison of model building architectures, various MT benchmarks (HellaSwag, HumanEval), various LLM studios (H20, LAMINI) offering such services, and ethics and trustworthiness aspects of it.
Python | H20 LLM Studio | LAMINI | Neptune | Arize | Hugging Face | Langchain | Stable LM | Fairness in NLP
Python | H20 LLM Studio | LAMINI | Neptune | Arize | Hugging Face | Langchain | Stable LM | Fairness in NLP
06/2022
-
12/2022
Research Data Scientist
MAX PLANCK INST. FOR PLASMA PHYSICS
(Auditing, taxes and law, 500-1000 employees)
Automated Scientific Discovery – Use of ML methods to ease the cycle of scientific discovery. Implemented an automated ML platform with several explainable AI and causality frameworks.
03/2022
-
05/2022
Full Stack Data Scientist
IU International University of Applied Sciences
(Internet and Information Technology, 1000-5000 employees)
Development of a feature-store to help boost model development process. Automation of machine learning models using the complete MLOps stack of AWS including CI/CD, model pipelining, deployment, and scheduling.
ETL scripts to load data google analytics, flatten the key/value structure of the data, and finally save it in AWS S3. Built a model for text labeling to better understand student feedbacks.
Tools use: Dask | AWS IAM | AWS MLOps | AWS CI/CD | Sagemaker Pipeline| Sagemaker Feature-store| PostgreSQL | S3 | SpaCy | NLTKr clients in materializing data-driven decisions by making use of fair and explainable machine learning models.
ETL scripts to load data google analytics, flatten the key/value structure of the data, and finally save it in AWS S3. Built a model for text labeling to better understand student feedbacks.
Tools use: Dask | AWS IAM | AWS MLOps | AWS CI/CD | Sagemaker Pipeline| Sagemaker Feature-store| PostgreSQL | S3 | SpaCy | NLTKr clients in materializing data-driven decisions by making use of fair and explainable machine learning models.
03/2017
-
12/2021
Lead Data Scientist
T-Mobile GmbH, Vienna, Austria
(Internet and Information Technology, 1000-5000 employees)
I have more than 9 years of experience in machine learning and data science. During my employments, I have lead teams of
several data scientists, both near- and off-shore and have built use cases delivering several hundred thousand € profit, purely based on data science and ML based solutions. I have ample experience of interacting with C-level executives on progress updates and strategy road maps. I have also led a group of senior researchers from universities on several projects about model bias, privacy, and explainable AI.
several data scientists, both near- and off-shore and have built use cases delivering several hundred thousand € profit, purely based on data science and ML based solutions. I have ample experience of interacting with C-level executives on progress updates and strategy road maps. I have also led a group of senior researchers from universities on several projects about model bias, privacy, and explainable AI.
02/2016
-
07/2017
Application Engineer
Samsung SDI Graz, Austria
(Industry and mechanical engineering, 500-1000 employees)
05/2012
-
02/2016
Researcher
Alpen Adria University, Klagenfurt Austria
(Internet and Information Technology, 500-1000 employees)
Worked on my PhD and publish the PhD thesis titled "Evaluation of Time Synchronization in Real Deployments of WSN".
Did thorough modeling on time-series data and used complext analytical tools such as state-space modeling and Kalman filter.
Did thorough modeling on time-series data and used complext analytical tools such as state-space modeling and Kalman filter.
09/2009
-
12/2010
Software Engineer
Ericsson Eurolab, kohlscheid, Germany
(Telecommunications, 1000-5000 employees)
02/2008
-
07/2008
Software Engineer
TRG (PVT) LTD, Lahore, Pakistan
(Internet and Information Technology, 500-1000 employees)
03/2007
-
12/2007
Software Engineer
WATEEN Telecom LTD, Lahore, Pakistan
(Internet and Information Technology, 500-1000 employees)
Certifications
Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
2018
Local Availability
Open to travel worldwide
Reisebereitschaft ist je nach Abstimmung vereinbart.
