Sahithi Akella

Sahithi Akella

Research Assistant

Followers of Sahithi Akella546 followers
location of Sahithi AkellaMunique, Baviera, Alemanha

Connect with Sahithi Akella to Send Message

Connect

Connect with Sahithi Akella to Send Message

Connect
  • Timeline

  • About me

    Product Owner - AI & Application Development

  • Education

    • Amrita University, Amritapuri Campus

      2010 - 2014
      Bachelor's degree Electrical, Electronics and Communications Engineering
    • Technische Universität München

      2014 - 2014
      Bachelor's degree Electrical and Electronics Engineering
    • Ulm University

      2015 - 2017
      Master's degree Electrical, Electronics and Communications Engineering
  • Experience

    • Ulm University

      Oct 2015 - Mar 2016
      Research Assistant

      Worked on different learning vector quantization algorithms

    • Nuance Communications

      Mar 2016 - Mar 2017
      Data Engineer - NLP

      Querying a database server through an API, post process, cleaning and ranking content using Python.Collecting text data from different websites of multiple languages through web scraping using PythonRunning tests based on Wikipedia-derived data sets.Process data for the language models in the Natural Language Understanding team.

    • Ulm University

      Aug 2016 - May 2017
      Master Thesis

      Estimation of the user’s cognitive load utilizing audio-based deep neural network architectures and learning strategies :Analyzing speech signals of the user for the classification of emotions. Working with artificial neural networks, pattern recognition tools, particular deep neural networks,such as auto-encoder networks, restricted Boltzmann machine and deep belief nets approaches,these are designed and compared to shallow architectures using training functions like Levenberg-Marquardt, Bayesian regularization and scaled conjugate gradient of multi-layer neural networks. The speech data is pre-analyzed and relevant spectral features (LPC, MFCC, ModSpec, PLP) are extracted. These extracted features serve as inputs to the classifiers. Results of the classifiers are collected andcompared by using statistical protocols (cross-validation and leave-one-subject-out) Show less

    • HARMAN International / Samsung Bixby

      Nov 2017 - Oct 2021
      NLP Engineer

      Developed chat-bot for the Samsung's Bixby voice assistant in German as part of a multi country Agile development team. Data collection through web-scraping, crowd sourcing, and linguistic tools to train ML models. Worked with named-entity recognition, dialogue systems, text and relation classification and functional semantics. Processed and analyzed large amounts of structured and unstructured data. Help set up data pipelines for usage data storage and processing and use it to improve system quality incrementally. Developed various web-tools using Google Web Toolkit and Java to work with SQL databases for linguisits to improve our datasets. Developed various web-tools using Google Web Toolkit with JAVA and SQL databases. Developed code generators and scrapers in Java to integrate SQL databases and data pipelines. Show less

    • Fraunhofer IAIS

      Nov 2021 - May 2023
      Research Engineer

      - Build language models for various media and medical related German and English data.- Developed a chatbot for medical domain using the RASA tool and worked with different extractors for entity extraction like Duckling and Spacy. - Working on text classification using deep learning and transformer models, mainly for news related data.

    • Tech Mahindra

      Nov 2022 - now

      • Gathered and analyzed requirements from stakeholders and team members for building and deploying a production application.• Prioritized requirements based on business value and technical feasibility.• Gained a comprehensive understanding of the technology stack involved in the application's development, including Next.js, TypeScript, and React.• Collaborated hands-on with the development team throughout the application development and deployment lifecycle, including deployment in Azure Containers.• Facilitated effective communication and collaboration among all stakeholders.• Planned and led sprint planning and analysis meetings. Show less - Designed and developed a robust text-based chatbot utilizing Python and relevant frameworks andintegrated with WhatsApp.- Hands on experience in Kubernetes and Azure to efficiently retrieve logs and monitor pods.- Worked with the operations team and engaged with various cross-functional teams to troubleshootand resolve production environment issues efficiently.- Managed a dynamic team responsible for task organization and allocation, ensuring the team’sefficiency in handling incoming tasks and meeting project deadlines.- Facilitated smooth and continuous communication between the team and other component teams,fostering an environment of collaboration and efficient functionality of the product. Show less

      • Product Owner - AI & Application Development

        Mar 2024 - now
      • AI Tech lead - Chatbot development

        Nov 2022 - Apr 2025
  • Licenses & Certifications