Manjula Mishra Email & Phone Number

Wayne, Pennsylvania, US

Wayne, Pennsylvania, US

• Lead high stake AI/LLMs based data science projects from ideation to production.
• Research new and emerging open-source as well as enterprise LLMs such as Gemini, Mistral, Llama, and Anthropic models to utilize them in real estate industry.
• Implement appropriate multi-modal LLMs model(s) with NLP and computer vision capabilities to consolidate structured real estate data and associated image data into useful products to build an intuitive property search.
• Design ground truth data collection, define success criteria, evaluation metrics, as well as dashboarding to collect feedback from QA.
• Create practical roadmaps for projects with realistic goals and phases to stay focused; present it to the leadership and collect feedback.
• Guide colleagues and junior data scientists new to deep learning to help them acquire new skills and finish their projects successfully.

Lehi, Utah, US

• Working on projects in NLP Conversational AI domain:
• Research and implement efficient and scalable solutions to automate intent labeling processes using open source solutions e.g. Snorkel.
• Explore and implement FAISS (Facebook AI Similarity Search) and other embeddings store options to optimize time for model training and other internal use cases.
• Create and deploy NLP solutions on AWS to QA structured and unstructured data using appropriate models such as KNN and K-means on pre-trained BERT and other available models closest to the conversation domain to help.
• Containerize models using Docker and deploy it on AWS to make Jupyter Notebooks UI accessible to NLU team for easy data manipulation.
• Collaborate with team members on publishing papers in NLP/multi-intent domain for continuous enrichment and knowledge sharing.

Berkeley, CA, US

• Extracted medical data from clients' EHR (Dr Chrono and Athena) APIs and store it on AWS MongoDB database for further use.
• Automated the data extraction, cleaning, and modeling process to ensure scalability and reproducibility.
• Cleaned and preprocessed raw text data in Python to generate insights, detect patterns, and use it for modeling.
• Built predictive models such as KNN, RF, and Keras to solve medical coding problems for doctors and medical coders.
• Containerized models using Docker to deploy them in staging and in production on AWS via APIs.Wrote custom metrics to analyze the performance of the models.

New York, NY, US

Mentored more than 50 students on data science topics.

San Francisco, California, US

• Perform data analysis on semi-structured data collected internally from Salesforce to understand and address the business needs of the company.
• Build custom resume parsers to handle unstructured text data from resumes.
• Analyze the integrated Salesforce data with text data to perform EDA using Pandas and NumPy libraries to find useful and relevant patterns in the data.
• Perform feature engineering to create new variable to capture the nuances of the text data.
• Apply various statistical and machine learning models including Logistic Regression, Random Forest to accurately predict the outcomes of the fellows in the program.
• Mentor 20-30 fellows on various data science topics to solidify their knowledge in ML, regression and classification techniques, and coding.

San Francisco, CA, US

• Served as teaching assistant for a core team of 9+ students, leading daily code challenges, live code review, and 1:1 feedback in Python.
• Provided daily feedback and issue resolution to 100+ Slack messages in help channels or code review.
• Led a team of 8-10 students to complete week-long cross-functional Build Projects, using the agile methodology.
• Led daily standup meetings to build camaraderie, facilitate sharing of ideas and work progress, and provide guidance for that week’s study material.
• Delivered supplemental resources on various Data Science topics, including Deep Learning and NLP.

Palo Alto, CA, US

• Conducted and summarized market research for the product fit.
• Researched and reported the competition in the same product space.
• Performed outreach to potential customers.
• Created an online database.
• Compiled and labeled the data in usable format.
• Conceptualized and designed company's initial web presence.

Port Jefferson, New York, US

• Reported to the CEO of Mission Be the effectiveness of mindfulness in the specific schools.
• Analyzed research data to measure the effectiveness of the Mission Be programs.
• Entered and Analyzed the data into the system to keep track of monthly/yearly budget.

Masters, Economics

Data Science