Custom Account Research

Last updated 6 months ago

Description: Custom Account Research enables users to specify an account specific research objective (i.e., GenAI initiatives) and use the Rox Agent Swarm to conduct research using both public and private data.
Demo

How are people using it today: Sales representatives are using Custom Account Research to streamline their account research process. Instead of manually conducting the same research on an account every few days, they can delegate this task to the Rox Agent. The Rox Agent has context about the products the sales representative is selling and the historical background of the account, allowing it to perform research tailored to their specific needs.

Custom Account Research enables sales representatives to automate their repetitive account research tasks
The best sellers are already using a smattering of ad-hoc generative AI tools such as ChatGPT to do research, but still have to manually queue up the research task everytime they engage.
Rox's Custom Account Research takes it a step further by providing personalized, context-aware research that considers your specific needs, your organization, and the account you're selling to. It also maintains provenance on previous research outputs, ensuring continuity and relevance.

Feature desiderata: fast (low-latency query response), transparent (show data sources), multi-modal (research over multiple data types).
Technical details

Architecture:
- Query Processing: recieves research objective -> decomposes query into atomic research queries over both internal SOR and public sources
- Data Retrieval: retrieves data over both private and public data (in SOR) and web data
- Research & Reasoning: Reasons over retrieved data, generating a contextualized final answer
Considerations:
- Speed: Utilizes hybrid search for optimal retrieval speeds, uses techniques such as parallelism, caching and batching
- Traceability: Track and store data sources to surface to user in the final answer.

Shareability: Account research will be exportable to various formats and applications, making it easier to share insights across your organization and integrate them into your workflow.
User -> Agent Driven Research: The Rox Agent, with its full context awareness, will proactively provide "Recommended Research" tailored to your needs.

Last updated 6 months ago

Description: Custom Account Research enables users to specify an account specific research objective (i.e., GenAI initiatives) and use the Rox Agent Swarm to conduct research using both public and private data.
Demo

How are people using it today: Sales representatives are using Custom Account Research to streamline their account research process. Instead of manually conducting the same research on an account every few days, they can delegate this task to the Rox Agent. The Rox Agent has context about the products the sales representative is selling and the historical background of the account, allowing it to perform research tailored to their specific needs.

Custom Account Research enables sales representatives to automate their repetitive account research tasks
The best sellers are already using a smattering of ad-hoc generative AI tools such as ChatGPT to do research, but still have to manually queue up the research task everytime they engage.
Rox's Custom Account Research takes it a step further by providing personalized, context-aware research that considers your specific needs, your organization, and the account you're selling to. It also maintains provenance on previous research outputs, ensuring continuity and relevance.

Feature desiderata: fast (low-latency query response), transparent (show data sources), multi-modal (research over multiple data types).
Technical details

Architecture:
- Query Processing: recieves research objective -> decomposes query into atomic research queries over both internal SOR and public sources
- Data Retrieval: retrieves data over both private and public data (in SOR) and web data
- Research & Reasoning: Reasons over retrieved data, generating a contextualized final answer
Considerations:
- Speed: Utilizes hybrid search for optimal retrieval speeds, uses techniques such as parallelism, caching and batching
- Traceability: Track and store data sources to surface to user in the final answer.

Shareability: Account research will be exportable to various formats and applications, making it easier to share insights across your organization and integrate them into your workflow.
User -> Agent Driven Research: The Rox Agent, with its full context awareness, will proactively provide "Recommended Research" tailored to your needs.