Data Annotation Services

We have an in-depth understanding of financial markets, accelerating your time to delivery.

Companies spend five times more on annotating data internally as they do with 3rd parties. And it will take your team members away from tasks where they can really add value.

You can fast track all your AI projects with bespoke labelled data sets. Get rapid, rigorous, and transparent results with real-time monitoring.

 

Annotation Services for Financial Institutions

Your data is your moat. Make it wider.

 
 

Equities

Annotate any topic or sentiment within corporate documentation such as annual reports and 10Ks, together with news and social media. Compile training data to leverage your in-house repository of equity research notes, investment committee decisions, and meeting notes. Build data to extract conclusions from academic research relevant to your to your investment portfolio. Create a repository of data for your investable universe based on industry reports, and specialist articles .

Fixed Income

Create annotated data sets for government bond, corporate bond and structured finance markets. Annotate key bond terms such as restricted entities, change of control, redemption options, and events of default. Work with All Street to build training data for bond covenants such as indebtedness, dividends, negative pledges, asset disposals and guarantee terms. For global macro, create data sets of central bank transcripts including sentiment of announcements.

ESG

Annotation as a Service powers your AI interrogation of ESG data lakes. Create categorised data sets based on the major frameworks including UN SDGs, UNPRI, GRI, SASB, EU Taxonomy. We help you annotate data from news, annual reports, sustainability reports, and specialist websites. Put together datasets of technical governance language from corporate policy or proxy voting documents. .

 

Back & Middle Office

Our annotation services help you create data sets for key back and middle office functions, including trade confirmations, payment instructions, trade finance and collateral agreements, custodian and prime brokerage agreements, and execution documents. Add value to front office automation initiatives by annotating documents produced daily such as research reports and market commentary. .

Legal

Legal annotation projects require not just domain knowledge but specialist drafting experience to capture legal nuance. At All Street, our legal annotation project leaders have wide ranging expertise in drafting capital markets documentation including (but not limited to !) ISDA Masters and confirmations, termsheets, LIBOR provisions, investment memoranda and prospectuses, MTN and other structured finance documentation.

Compliance

Your processes may involve particular types of compliance documentation that are not covered by RegTech providers and you want to create your own models for fraud detection, KYC/AML, and interrogation of financial regulations. We work with you to create robust annotation rules designed to minimise false positives and false negatives for anomaly detection. Create datasets of named entities relevant to your business including people, companies and locations.

 

End to End Annotation Process

1. Define your task

It is of paramount importance to correctly specify the problem to be solved. We can assist you with preparation of a domain ontology, which is a representation of the types, properties and inter-relationships of the entities which machine will learn about. For example, what information should be labelled as “Climate Action” and what should go under “Affordable and Clean Energy” when classifying the UN Sustainable Development Goals.  This will ensure that your final AI algorithm communicates its results in a consistent way.

2. Upload Data

If you have a comprehensive internal data corpus you can upload it directly to our secure platform. Alternatively we can create a dataset for you as per your requirements. Our annotation systems handle all major file formats including PDF, HTML, Word, Powerpoint, Excel, CSV, and JSON.

3. Review progress

Clients using All Street’s Annotation as a Service have their data labelled by the same pool of professional analysts that label data for the SEVVA AI platform. The analysts are experienced in understanding complex financial and legal documents and making nuanced technical determinations.  You can monitor progress in real time including annotator consensus and projected completion time via the tracking dashboard.

4. Download results

Rapid results in a matter of hours or days. Download the results and use them to build your own machine learning models. We can also work with you to design and deliver your models, including building customised Cognitive Robots which you can access in Sevva.

Flexible Annotation Tasks

 

Text Classification

Sentence, paragraph or document level.

Sentiment Classification

Positive, negative and others such as complexity and uncertainty.

Data Extraction

Tagging of keywords and named entities.

Sentence Structure

Including part of speech, semantics, and morphology.

 
 

Supercharge your data lake

 

Annotation as a Service by All Street. Contact us today to discuss your annotation project and get a free quotation.