Processing of Legal documents

Labeled data generation and use of Python libraries to built data pipeline by data pre-processing / cleaning

Building supervised machine learning algorithms to identify risks from legal documents and classify those as low / medium / high.

Feature Engineering using NLP; Tuning of Hyper parameters and ensemble model to improve accuracy.

Semi-automation of training data generation, comparison of models to decide suitable model for deployment.

Specific Data Extraction from Scanned Documents

Extraction of complete data from scanned documents.

Use of various NLP and ML techniques to extract specific values from the unstructured data

Elastic search with NLP to store the specific data and make the data searchable

Build APIs to be integrated with other application

Use of ML in Healthcare

Use of python libraries for data exploration; to built data pipeline by data pre-processing

Use of ML algorithms to get baseline accuracy

Feature Engineering using encoding, label power set and multi-label classification

Verification using Cross validation accuracy

Model deployment, availability through API and integration with platform

Best Strategy to recover Dues from Defaulters

Labeled data generation and use of Python libraries to built data pipeline by data pre-processing / cleaning

Identification of independent parameters for ML based on

a) 360 degree view of the customer profile;

b) Historical data from collections systems (action-result) and the performance data of the collectors

Built ML models on the parameters to predict best strategy, best possible action string and suitable collector / agency.

SMART BOTs for Finance, Healthcare, Project Management and Recruitment Industries:


Built voice bot for a leading global service provider for booking doctor’s appointment in a faster way.


Built SMART BOT to help our client to close the recruitment positions quicker there by helping the business to increase the revenue in lower cost.


Our client is a leading technology company in banking domain. bizAmica has built SMART BOTs to help their clients. Text and voice enabled bot creates differentiated offering for the banks during various loan inquiries.

Another interesting SMART BOT is to get “promise to pay” from loan installment defaulter by negotiating with the defaulter.

Exhibitions and Events Industry

Built SMART BOT for exhibition and events industry. BOT helps event organizers to automate repetitive tasks and to save precious resources on event management. The BOT helps end users pre-event, during event and post-event engagements.

Data Science for logistics domain:

Client is a global leader in Big Data and Analytics space.

Identification of independent parameters influencing the return of consignments

Data pre-processing of 1.15 M records.

Use of various ML algorithms (Decision Tree, Random Forest, Naïve Bayes, KNN, Gradient Boost)


Building XAI Platform for Healthcare

  • High level architecture design and detailed low level designs for individual modules
  • Research, innovation and implementation of human collaboration aspect
  • Implementation and integration of modules in agile methodology
  • Explanatory Model Debugging
  • Validation on use-cases from healthcare sector
    • Explaining the Opioid Risk predicted for a patient based on his past medication, therapies, etc. and recommending the alternate procedures to minimize OUD risk
    • Explaining patients medical claims when predicted for rework with 90%+ accuracy using  TensorFlow DL models
  • ML modelling on use-case datasets and generating human readable explanations with recommendations
    • ML models to predict Opioid Risk (High/Medium/Low) based on patients’ past records
    • ML models to predict whether a patient’s claim has high chances of Rework with 90%+ accuracy  using  TensorFlow DL models

Techno-commercial solution in XAI (Explainable AI):

  • Researched current state, developments and limitations in Explainable AI
  • Defined innovations, R & D objectives and milestones
  • Solution architecture and technical scalable architecture
  • Building various modules of the platform and integration using Agile methodology
Recruitment Bot