Private Context Engine API details


The Selerity Private Context Engine API (PCE) automatically generates metadata related to client interests and requests for quotes, identification of specific instruments, quoted prices, and trade execution details from unstructured chat messages.

The metadata can be used to automate enterprise workflows including customer relationship management, compliance, regulatory reporting, research distribution, behavioral analytics, and market intelligence.

PCE runs as a deployed application instead of Software-as-a-Service (SaaS), highly sensitive electronic communications do not leave the client’s environment. PCE can be applied to capital markets chat communications and other unstructured data types such as email, research, voice, news, social media.

PCE consists of two components:

Data Extraction

Automated extraction of instruments, trades, subjects, financial concepts from chats and emails.

Analytics and Insights

Automated insights generated from extracted data to facilitate increased trading revenues and productivity.

PCE automatically captures client interests from digital communications. It allows sales and trading to understand better their clients’ needs, and facilitates predictive trade matching and client relationship analytics.

PCE can parse InstantBloomberg chat messages, Symphony chat messages, emails, and other unstructured digital communications with clients.

PCE automatically identifies trade-related conversations from large pools of unstructured messages, many of which are low relevance, and converts unstructured conversations into machine-readable trade-related metadata like instrument, price, size, and direction.

Analyzed data is available via PCE Analytics API which you can use to query, among other use cases:

  • Historical expressions of clients’ interests.
  • Find counterparties for current clients’ interests.
  • Use historical price interests dataset for your price modeling.
Corporate Bonds
Archive size
Archive history
Information source
Bond trader chats
Content type
Bond trader chats
Delivery protocol
Authentication type
Token Authentication
Response format
Update frequency
Point in time

Frequently asked questions

Can PCE manage Bloomberg’s Instant Bloomberg archive?

Yes.  PCE has built in parser can analyze Instant Bloomberg archive.

What are the data fields that PCE can extract from conversations with clients?

PCE extracts over 28 different fields and over 100 different values.  It includes ticker, coupon, maturity, price, type of price, size, client direction, dealer directions, and etc.

Can PCE parse messages from muti-participant chat room?

Yes.  PCE can analyze complex conversation in multi-participant chat rooms. PCE identifies an instrument of the conversation, and follows all the conversation involves the instruments. 

Privacy is important to us, can PCE run within a bank’s on-prem?

Yes. PCE can be deployed to a bank’s on-prem.  PCE does not required any outside network connectivity to process the bank’s chat messages and extracted data.