Operational Efficiency
How an AI-powered CV search engine matches experts with public tenders
Responding to public sector tenders requires identifying the right experts fast, and getting it wrong means lost opportunities. Agilytic partnered with a European consulting firm to build an AI-powered CV search engine, turning a corpus of over 25,000 profiles into a strategic asset for tender response.
To protect confidentiality, we may alter specific details while preserving the accuracy of our core contribution.
Context and objectives
A European consulting firm needed to quickly identify qualified experts from a database of thousands of CVs to respond to public sector tenders.
The existing process relied on manual, full-text searches that were:
Slow and inconsistent
Dependent on individual knowledge rather than data-driven criteria
Prone to missed tender opportunities
The client required an on-premise solution to comply with strict security constraints and integrate with legacy infrastructure. The objective was clear: build a CV search engine that could surface the most relevant profiles reliably and at scale.
Approach
Defining selection criteria with business teams
Agilytic collaborated with the client's business teams to identify the signals that matter most when shortlisting experts. These included:
Years of experience
Leadership background
Language skills
Education
Domains of expertise
Structuring CV data at scale
An LLM-based extraction pipeline was built to process long, unstructured CVs and convert them into structured data profiles. This step ensured consistent, machine-readable information across the full corpus.
Combining filtering and semantic ranking
The CV search engine uses a two-stage approach:
Hard-criteria filtering to narrow down candidates based on defined requirements
Semantic ranking to surface the most relevant profiles based on contextual relevance
A simple interface was delivered for business users to triage results. The solution was deployed in line with security and infrastructure constraints.
Results
The deployed CV search engine directly addressed the client's core challenge of identifying qualified experts for public tenders from a large and growing database.
Key outcomes included:
Significant reduction in search time, replacing manual, time-consuming processes with an efficient automated system
Lower risk of missed tender opportunities, thanks to broader and more consistent coverage of the CV corpus
A shift from personal knowledge to data-driven expert selection, enabling more objective and reliable shortlisting
Business user feedback confirmed strong adoption and satisfaction with the tool's speed and relevance.
To safeguard confidentiality, we may modify certain details within our case studies.