Comprehensive digital solutions designed for impact and scalability.
Domain-driven solutions tailored to the unique needs of each industry.
Technical excellence backed by years of hands-on experience.
A journey into innovation, culture, and customer success
The client is a Switzerland‑based clinical research and competitive intelligence firm serving pharma, biotech, and medtech. It performs expert‑led reviews of scientific literature and clinical studies, conducts primary interviews, and validates findings against trusted sources. The firm tracks pipelines and research activity to inform product and market decisions. With analysts active in 50+ countries, it delivers concise, credible findings for executive teams in R&D, medical, regulatory, and market access.
Country
Switzerland
Industry
Healthcare
Services Used
Researchers at the clinical research firm spent significant time on manual tasks while gathering scientific literature from trusted sources such as PubMed, SciDirect, MedRxiv, BioRxiv, or using Boolean searches on Google. Articles frequently appeared across multiple databases, requiring tedious manual deduplication. Each researcher devoted 20–30 hours per week to repetitive processes like searching, filtering, ranking, categorizing, and summarizing publications. Preparing newsletters and research briefs for pharmaceutical clients added to the workload, often slowing delivery timelines.
To overcome these challenges, the firm envisioned an AI-assisted research platform that could replicate existing workflows while automating repetitive steps. At its core, the platform would serve as a unified interface, eliminating the need for researchers to check multiple databases separately. Instead, all literature review tasks could be streamlined within one environment, with researchers maintaining an oversight (human-in-the-loop) approach to validate outputs and ensure accuracy.
To bring this vision to life, the organization partnered with Daffodil Software to build a tailored AI-assisted research solution.
The following were the key requirements discussed:
Build a single platform that consolidates articles from multiple scientific databases, eliminating the need for researchers to run the same queries across different sources manually.
Connect directly with trusted databases such as PubMed, MedRxiv, and ScienceDirect through their official APIs, ensuring structured, accurate, and compliant access to scientific literature.
Incorporate intelligent mechanisms to detect and remove duplicate records when the same article appears in multiple sources, delivering cleaner, more reliable datasets.
Reduce manual effort by automating routine research activities such as advanced searching, relevance ranking, categorization into themes, and generating concise article summaries.
Give researchers the flexibility to package results into multiple, client‑ready formats - structured spreadsheets, professional PDF reports, or newsletter‑style deliverables.
Design the platform as an AI‑assisted research tool that automates repetitive tasks while keeping researchers in control to review, validate, and finalize results.
Include functionality to monitor pre‑published studies and detect when they are updated or finalized, with changes reflected automatically in reports and deliverables.
Daffodil helped the client by developing an AI-assisted research platform tailored for clinical and pharmaceutical researchers. The platform integrates directly with six leading scientific databases – PubMed, MedRxiv, BioRxiv, ScienceDirect, MDPI, and The Lancet – and delivers consolidated results in a single interface. This replaced the manual process of running the same search on multiple sites and removing duplicate articles by hand.
To ensure accuracy while reducing repetitive work, the platform was built with a Human-in-the-Loop AI design. The system takes over routine steps such as removing duplicate entries, ranking articles for relevance, organizing them into categories, and generating structured summaries. Researchers still review, validate, and interpret the findings themselves, keeping the final judgment entirely in their hands.
The platform was designed with these core features:
A single structured query is executed across connected scientific sources in parallel via APIs, and results are returned in one interface with a consistent schema (title, authors, DOI, abstract, source, publication date). Source attribution is retained for every record, and query-time filters (date range, source selection) are supported to narrow results at ingestion.
Records are matched and merged across sources using identifiers (DOI, title similarity, author/date), removing duplicates before review. Each article is scored as High, Medium, or Low relevance using LLM-assisted abstract-to-query alignment, promoting the most pertinent studies for analyst attention while minimizing noise.
Grid View provides a spreadsheet-style layout with raw article data, metadata, filters, relevance tags, and on-demand extraction of objectives and conclusions for precise manual validation.
AI-Assisted View generates concise summaries by default using LLMs, surfacing study objectives, findings, and conclusions to speed comparative analysis without replacing expert judgment.
Articles are automatically classified into research themes to organize literature by topic and streamline downstream reporting. Pre-published records are tracked for status changes and revisions; when updates occur, the system flags changes and routes them to the appropriate section in recurring outputs (such as newsletters), preventing redundancy.
Reviewed results can be exported as consolidated grids, structured reports, or newsletters. Exports preserve citations, relevance markers, categories, and update flags, enabling consistent, client-ready deliverables without additional formatting work.
The AI‑assisted platform transformed how researchers conducted medical research by cutting repetitive tasks such as searching, filtering, and summarizing. Weekly effort dropped from 20–30 hours to just 4–8, while automated deduplication and relevance scoring reduced noise by 60–70%. With six leading databases unified into one interface, researchers no longer needed to repeat queries across multiple sources, ensuring comprehensive coverage without duplication.
The system also raised the quality of client deliverables. Research updates could be exported in minutes as structured reports, with appropriate citations, categories, and relevance scores. This balanced approach of speed, accuracy, and oversight has positioned the client as a leader in medical research.
Developing an AI-powered telehealth application for a US-based health tech firm
Enhancing development capabilities for a premier fitness app company through team augmentation services
Developing a feature-rich Gen-AI platform for the UK’s leading global investor in sustainable infrastructure.
Let us understand your business objectives, set up initial milestones, and plan your software project.
Δ