Ashish Lal Email & Phone Number

Mumbai, Maharashtra, India

An EXL company

United States

Using gpt index to add voice transcripts, FAQs, website to ChatGPT for query.Using pinecone to index and query data and using OpenAI to conver onto material language.

Mumbai, Maharashtra, India

NLP Analysis using transformers - Prediction of a job seeker's career path given current role, experience, technical skills.Analyzing course transcript, description, syllabus for whether the transcript adheres to syllabus, title keyword extraction, grammatica error correctionExploration of reinforcement learning in recommender systems - contextual bandit, SlateQ algorithms.

Thane, Maharashtra, India

1. Role of Data Scientist with Big Data Team2. Creating Reports, charts, analyzing company datasets.3. Creating healthcare models and models for human facial emotion.4. Time series models, ARIMA for new user forecasting 5. Azure Data Factory, Database migration action from AWS to on prem6. Chatbot POC using RASA for booking appointments

Mumbai Area, India

Finacus SolutionsData Science Consultant/Program ManagerLead Finacus AI team1. Design, architect and drive the development of AIframework at Finacus to be used in building models. Abanking model was built for identifying prospectivecustomers who would take a loan in future. Keywords:LightGBM, XGBoost, Keras, Neural Networks, RandomForest2. Drive the banking chatbot development usingconversational AI – Rasa framework was used to build abanking… Show more Finacus SolutionsData Science Consultant/Program ManagerLead Finacus AI team1. Design, architect and drive the development of AIframework at Finacus to be used in building models. Abanking model was built for identifying prospectivecustomers who would take a loan in future. Keywords:LightGBM, XGBoost, Keras, Neural Networks, RandomForest2. Drive the banking chatbot development usingconversational AI – Rasa framework was used to build abanking assistant.Software Design and Development (in VoIP area), Consultant in Java programming, C++, C# Clients -1. Infinet Financial Systems (now Enepath), Mumbai , INDIA: VoIP based trading system in C++2. ButterfllyMX, NY, USA: VoIP based video calling system in C#3. Blobcity, Mumbai, India - Java related development Show less

Having fun at Kaggle Competitions1. Use Case: Lung Cancer Detection using deep learningObjective: Using a data set of thousands of high-resolution lung scans provided by the National Cancer Institute, USA, and Kaggle, develop algorithms that accurately determine when lesions in the lungs are cancerous. Methodology: The solution was implemented using Keras, U-NET architecture for segmentation and 3d CNNs (convolutional neural networks). The CNNs and UNET were implemented in Keras… Show more Having fun at Kaggle Competitions1. Use Case: Lung Cancer Detection using deep learningObjective: Using a data set of thousands of high-resolution lung scans provided by the National Cancer Institute, USA, and Kaggle, develop algorithms that accurately determine when lesions in the lungs are cancerous. Methodology: The solution was implemented using Keras, U-NET architecture for segmentation and 3d CNNs (convolutional neural networks). The CNNs and UNET were implemented in Keras and pretrained models were used. Other tools used were numpy, pandas, other python libraries for DICOM images.All the models were trained and tested using a GTX 1080 Ti 11 GB GPU.2. Use Case: Histopathologic Cancer detection.Objective: Use deep learning to identify metstatic cancer in small image patches taken from large digital pathology scans.Methodology: The solution was implemented pytorch. A set of pretrained CNN models (resnet, densenet, squeezenet etc) was trained on different image sizes and an ensemble was taken for the final prediction.Tools used: Keras, pytorch, numpy, pandas, scikit-learn.3. Use Case: Personalized medicine: redefining cancer treatmentObjective: Use deep learning and machine learning to predict the effect of genetric variants to enable personalized medicine.A neural network model written in Keras was also used to predict genetic variant classes.4. Use Case: Human Protein Image Classification using deep learningObjective: Classify mixed patterns of proteins in microscopic images. Analysis: Develop deep learning models capable of classifying mixed patterns of proteins in microscope images. Methodology: The solution was implemented in Keras and pytorch. A pretrained CNN was used to identify proteins in images. The CNN was then trained over multiple image sizes. Then an ensemble of models was created to produce the final prediction.5. Use Case: Find the nuclei in divergent images to advance medical discovery. Show less

Mumbai, Maharashtra, India

1. Manage the Finacus AI team.2. Provide technical leadership including transfer oftechnology to AI team members.3. Design, architect and drive the development of AIframework at Finacus to be used in building models. Abanking model was built for identifying prospectivecustomers who would take a loan in future. Keywords:LightGBM, XGBoost, Keras, Neural Networks, RandomForest4. Drive the banking chatbot development usingconversational AI – Rasa framework… Show more 1. Manage the Finacus AI team.2. Provide technical leadership including transfer oftechnology to AI team members.3. Design, architect and drive the development of AIframework at Finacus to be used in building models. Abanking model was built for identifying prospectivecustomers who would take a loan in future. Keywords:LightGBM, XGBoost, Keras, Neural Networks, RandomForest4. Drive the banking chatbot development usingconversational AI – Rasa framework was used to build abanking assistant.5. Developing a facial recognition system for social mediamonitor system. Also extensively used Google cloud APIfor landmark detection in images, speech to texttranscription and translation.6. Provide technical leadership including transfer oftechnology to AI team members.7. Be hands on in python coding, feature engineering andmachine learning, neural networks.8. Identify and scope data science projects and create POCswhen required. Show less

Thane, Maharashtra, India

Taught machine learning to student’s enrolled in ITVEDANT’smachine learning course.

Mumbai Area, India

Embedded software design and development.

San Francisco Bay Area

Softswitch support, feature development

Greater New York City Area

Wifi, WLAN related firmware development

Greater Boston Area

R&D, test tool development

Master Of Science, Computer Science

GPA 3.93/4.0

Bachelor Of Technology (B.Tech.), Computer Science And Engineering, Gpa 9.21/10.0

AIR 809 JEE 1990,