Emerging Trends/Opportunities Identification: 16 weeks Implementation

Tiger Analytics' Emerging Trends Identification Solution, powered by Azure, enables organizations to identify emerging trends within a target category, quantify the potential business opportunity linked to those trends, and pinpoint key growth drivers to capture the opportunity. This is achieved by leveraging Azure AI and Data Services to triangulate consumer insights from social media, product reviews, search behavior, and research data, along with internal sales and marketing data.

At the core of the solution is a proprietary pattern-matching algorithm that analyzes and compares multiple time-series data patterns across different timeframes to accurately identify and assess the value of emerging trends. Key use cases supported by this Azure-based solution include New Product Introduction and Mergers & Acquisitions, where organizations gain data-driven insights to guide strategic decisions.

The solution is completely cloud native and works seamlessly with Azure PaaS components. Listed below are the key steps in the workflow

1. Data is ingested from various sources like Point of sales systems, Search engines, Social media providers and other 3rd party data sources (Netbases, Semrush, sprinklr etc). Data is partitioned and placed in ADLS Gen 2. Azure Data Factory is used for data ingestion
2. Data is transformed using ADF/Databricks and persisted in ADLS Gen 2
3. ML models are built and trained in custom Python and Azure Machine Learning using custom algorithms available as part of Tiger Emerging Trends solution. In the first step NLP models are used to identify trends of interest followed by pattern matching models to forecast the market potential of the trends.
4. Azure Databricks is used for computation
5. The ML Pipelines leverage Azure Functions for serverless compute and Azure Load Balancer for distributing traffic.
6. Access Control and authentication for data and the ML workspace leverages Azure Active Directory and Azure Key Vault.
7. Model performance is evaluated using log metrics and monitored using Azure monitor
8. Dashboards are provided for consumption using Power BI
9. Models are retrained as necessary in Azure Machine Learning

The solution comes with a business-friendly front end with a plethora of pre-built reports & dashboards to consume the findings: a) Dashboard showing key trends of interest with various KPIs to understand the relevance to the organization. b) Visualizations for “Most Likely Journey” of trends across different channels (social, product launch, search etc) c) Opportunity Analysis Dashboards: Trend life cycle chart over 5 year horizon d) Demand driver analysis charts