alva is now Penta. We are the world’s first comprehensive stakeholder solutions firm. Learn More

Learn More

Hit enter to search or ESC to close

Get in touch

Our model

alva singularly integrates Technology, Methodology and Expert Analysts to provide actionable intelligence for our clients.

Technology applies unique NLP capabilities applying lessons learned over many years specifically for reputation intelligence. We apply machine learning and semantic analysis to quantify sentiment in complex, long, and specialised types of content.

Methodology allows us to extract insight from millions of content items through different lenses including companies, geographies, influencers, issues, media types, products, and stakeholders.

Expert Analysts apply industry understanding to interpret findings, tailor analytics intimately to clients’ needs and integrate with internal KPIs, and connect findings to specific actions.

Step 01.

Data Gathering

Our proprietary technology aggregates a signification proportion of publicly available content including traditional media (print, broadcast, online), social media (twitter, blogs, forums), surveys and analyst reports-in real time, globally, 24/7.

Step 02.


Filters and semantic analysis shape the data, which is then analysed by machine learning models to identify and categorise company, issue, product, sector, stakeholder, and several other sets of entities and topics.

Step 03.


An early pioneer in natural language technology, alva developed the first ”long-form NLP” to score nuanced sentiment on complex documents. The system applies techniques based on many years of continual learning to understand specialised content types such as investment management research and regulatory filings.

Our technology understand large volumes of streaming content to identify companies and other entities, topics and themes, industry sector, stakeholders, geographic location, and several other dimensions.

Sentiment is calculated through a multi-step process incorporating rules and statistical NLP to shape data then applied to deep learning models.

Step 04.


The system places sentiment and topics into context to see emerging trends, benchmark against peer groups, and drivers that change stakeholder perceptions. To obtain “Stakeholder Intelligence”, alva integrates multiple internal and external datasets linked to clients’ business KPIs.

Step 05.

Decision making

Analysts apply industry understanding to connect analytic findings into specific insights and recommendations tailored to clients. Analysts carefully customise alva’s broad coverage to specific needs of individual clients.

Get in touch

and we will be happy to answer your questions.