Every inbound deal captured and scored the moment it lands

The problem. Deal flow landed in several partners’ inboxes, screening standards varied, and the CRM ran as a backlog.

1. 1Detected teasers and CIMs in forwarded emails
2. 2Extracted company, sector, revenue and EBITDA with page references
3. 3Scored against the firm’s thresholds and wrote the record to Salesforce

Source documents into a first draft, with judgment applied

The problem. IC memos were drafted by hand from CIMs, interviews and financials, taking days on time-boxed deals.

1. 1Pulled the key financials from the CIM and statements
2. 2Built revenue and EBITDA bridges, each figure linked to source
3. 3Marked every add-back as keep, adjust or flag, with the reason

A 248-page prospectus into a source-linked capital structure

The problem. Analysts read 200+ page prospectuses by hand to find the capital structure, triggers, reserves and hedging terms, then retyped them.

1. 1Recognised the asset class and applied the matching schema
2. 2Pulled out structure, ratings, coupons, triggers, reserves and hedging
3. 3Checked every field and linked it to the paragraph it came from

The right comparables for any CLO tranche, scored and traceable

The problem. Picking comps was manual workbook work, and which bonds counted as comparable was a judgment call that was hard to audit.

1. 1Filtered the universe on rating, manager tier, reset and non-call dates, MVOC
2. 2Scored the bonds that passed and compared each on its colour date
3. 3Linked every data point back to the query that produced it

Every originator’s data pack in one format, ready to compare

The problem. Each originator sends data in its own layout, so analysts re-mapped fields by hand before any comparison.

1. 1Extracted every table from each pack
2. 2Mapped it into one standard schema
3. 3Answered questions across all originators and vintages

What moved across 150+ names, and what it means for the book

The problem. Tracking 150+ names meant switching between Bloomberg, filings and research, and news reported events without saying what they meant for the loans held.

1. 1Grouped every name by sector and flagged the biggest movers
2. 2Explained each move with evidence from filings and research
3. 3Marked each as a direct hit or read-across, with a next step

One company scored on three lenses of AI disruption, every point cited

The problem. Every deal needs a view on AI risk, but it rested on opinion and a few slides.

1. 1Ran the outside-in assessment on the CDD report from our portal, with no management inputs
2. 2Scored external disruption on seven parameters, including how easily AI could replicate the offering
3. 3Scored internal readiness only on AI confirmed in use, and graded ten technology tracks