Case study · Professional services

Leadenhall Search: candidate feedback worth reading.

Fifty One Degrees built a proof of concept for Leadenhall Search, a financial services-focused search and recruitment firm, that turns the unstructured documents recruitment already produces, job descriptions, CVs and client feedback, into detailed, role-specific feedback for every candidate.

Leadenhall Search case study
51.4°N · 0.1°W
Client
Leadenhall Search
Sector
Search and recruitment (financial services)
What we built
A RAG-based candidate feedback generator
Built from
Job descriptions, CVs and client feedback
The challenge

Every candidate deserves feedback. No recruiter has the time.

In specialist search, the candidate experience is the brand: today's rejected candidate is tomorrow's client. Leadenhall Search held everything needed to explain a decision, the job description, the CV, and the client's own feedback on why a candidate did or did not progress, but that information lived in unstructured documents, and turning it into thoughtful individual feedback by hand did not scale.

What we built

Feedback grounded in the real documents

Reads the documents recruitment already produces

The solution works directly from unstructured inputs, job descriptions, candidate CVs and written client feedback, with no need to restructure the data first.

Retrieval augmented generation

RAG grounds every generated response in the actual documents from that search, so the feedback reflects the real role requirements and the real reasons for the decision, not a template.

Specific, usable candidate feedback

Instead of a generic rejection, candidates receive feedback that names the strengths and gaps the client actually saw, the kind of response that keeps a candidate warm for the next role.

How we built it

Proven on real searches

Real roles, real candidates

The proof of concept was built and tested on ten real candidate cases across risk and regulatory roles, credit risk, market risk, transaction reporting and data governance.

Client feedback as the source of truth

Where the client had given written feedback, the system used it as the backbone of the candidate response, so the output matched what was actually said.

Example-led quality control

Ideal feedback examples from Leadenhall's own consultants set the bar for tone and detail, keeping the generated responses in the firm's voice.

Fifty One Degrees showed that the documents a search firm already holds, job descriptions, CVs and client feedback, contain everything needed to give every candidate specific, useful feedback instead of a generic rejection.Fifty One Degrees x Leadenhall Search
FAQ
What did Fifty One Degrees build for Leadenhall Search?

A proof-of-concept solution using retrieval augmented generation that reads job descriptions, CVs and client feedback together and generates detailed, role-specific feedback for candidates, in place of generic rejection messages.

Why does automated candidate feedback matter for a search firm?

Because in specialist recruitment the candidate experience is the brand. A rejected candidate who receives specific, respectful feedback stays warm for future roles and often becomes a client. The information needed to write that feedback already exists in every search; the constraint is recruiter time.

Can Fifty One Degrees automate parts of my recruitment workflow?

Yes. Fifty One Degrees builds AI solutions on the unstructured data recruitment firms already hold, from candidate feedback to search support. Book a Discovery Call and we will find where AI pays back first in your workflow.

See all client work →
Next step

Want AI working on your firm's unstructured data?

Book a 30-minute Discovery Call and we'll find the highest-value use of the documents your business already produces.