WelfareWise.ai : Transforming European Welfare Narratives into Actionable Insights for Impact

AgentHack submission type

Enterprise Agents

Name

Nisarg Kadam

Team name

WelfareWise.ai

Team members

@danny_lim

How many agents do you use

Multiple agents

Industry category in which use case would best fit in (Select up to 2 industries)

Public sector

Complexity level

Advanced

Summary (abstract)

Revolutionizing social work diagnostics through intelligent automation to enhance decision-making and case management, our innovative solution builds on an earlier experimental agentic prototype and now harnesses more advanced UiPath technologies to streamline workflows and improve client outcomes — using the European welfare system, one of the world’s best, as a case study to deepen understanding of how agentic automation can be effectively deployed.

Detailed problem statement

In the field of social work, practitioners routinely collect large volumes of free-text notes during home visits, phone calls, and needs assessments. These notes, while rich in context, are often unstructured, verbose, inconsistent, and fragmented. This lack of standardization makes it extremely challenging to quickly extract and synthesize critical information about each client’s circumstances.

As a result, social workers face significant delays in identifying urgent needs and developing effective intervention plans. The manual effort required to review, interpret, and organize unstructured notes not only increases the administrative burden on already stretched professionals but also introduces the risk of human error, inconsistent decision-making, and important details being missed. These challenges can lead to slower response times for families in need, reduced service quality, and difficulty in ensuring compliance with reporting and audit requirements.

Ultimately, the inability to systematically process and analyze social work narratives limits the scalability, consistency, and impact of welfare programs, highlighting the urgent need for solutions that can transform unstructured data into actionable insights—empowering social workers to focus on what matters most: supporting vulnerable families effectively and efficiently.

Detailed solution

To address the challenges of unstructured and inconsistent social work notes, we present WelfareWise.ai—an intelligent, agentic automation platform designed specifically for social sector diagnostics and case management.

Our solution leverages advanced UiPath technologies—including agentic orchestration, AI-powered document understanding, and human-in-the-loop capabilities—to transform raw, unstructured narratives into structured, actionable insights across key social welfare domains. The automated workflow begins by ingesting family case notes and related documents from emails or other channels, using robots to capture and prepare the data for analysis.

WelfareWise.ai employs a layered architecture, inspired by modern agentic AI systems, to ensure clarity, scalability, and modularity:

Perception Layer:
Robots automate the collection of case notes and documents from various sources (email, cloud storage, forms), performing initial validation and extraction to ensure high-quality input.

Reasoning Layer:
Specialized agents—including the Orchestrator and Domain Assessment Agents—analyze, interpret, and structure the data. They split the case into major social welfare domains (Health, Housing, Income Security, Education, Family Functioning, Financial Resilience), then apply context-driven logic to extract strengths, risks, goals, and recommended actions for each.

Action Layer:
Outputs from the Reasoning Layer are consolidated and formatted into structured reports, case summaries, and recommendations—automatically archived, distributed, or forwarded to downstream systems as needed.

Feedback Layer:
Human-in-the-loop review steps allow social workers to validate, adjust, or supplement the AI-generated assessments and recommendations, ensuring that expert judgment remains central and that the automation continuously learns from human input.

Memory Layer:
All case interactions, decisions, and outcomes are logged and stored for audit, compliance, and future reference—enabling robust tracking, reporting, and continuous improvement.

This layered approach ensures each stage of the process—from data collection to reasoning, action, feedback, and memory—is transparent, configurable, and easy to extend or adapt to other welfare systems and countries.

With WelfareWise.ai, social services agencies can achieve:

Dramatic reductions in manual workload and administrative time (processing cases in minutes, not hours)

Improved accuracy, consistency, and compliance through standardized, explainable assessments and audit trails

Scalable support for thousands of families, adaptable to any welfare framework

Faster, more reliable interventions that put critical insights directly into the hands of decision-makers

By deploying WelfareWise.ai—demonstrated using the robust frameworks of the European welfare system—agencies can deliver smarter, fairer, and more efficient support to vulnerable families, transforming how social work diagnostics are performed at scale.

Demo Video

Expected impact of this automation

Expected Impact of WelfareWise.ai (European Context)

Massive Time Savings:
Cuts case processing time from 1 hour to under 5 minutes—saving up to 100,000 work hours/year for a typical EU country.

Significant ROI:
Frees up the equivalent of 45–50 full-time staff (worth €3.6M/year), enabling agencies to do more with the same resources.

Quality & Compliance:
Ensures consistent, rule-based case assessment and audit-ready logs—meeting strict EU welfare standards.

Better Service & Faster Help:
Families receive support in hours, not weeks; social workers focus on people, not paperwork.

Scalable & Adaptable:
Easily reconfigurable for any national framework, language, or policy—future-proofing your welfare agency.

UiPath products used (select up to 4 items)

UiPath Agent Builder
UiPath Apps
UiPath Integration Service
UiPath Maestro
UiPath Orchestrator
UiPath Studio Web

Integration with external technologies

OneDrive, Outlook 365, Slack, GenAI Connector

Agentic solution architecture (file size up to 4 MB)

Sample inputs and outputs for solution execution

Sample Input Description
Input Type:

Unstructured family case notes, assessments, or interview summaries—often in free-text format, collected during home visits, phone calls, or via digital forms.

Structured data files (Excel/CSV) containing multiple family attributes (e.g., housing status, health needs, income sources, education records, family functioning notes).

Sample Input Example:

Email with attachment: A social worker submits an email with a link to a OneDrive Excel file named family_case_notes.xlsx.

Excel/CSV columns might include:

Family ID
Interview Date
Health Notes
Housing Situation
Income and Employment Details
Education Notes
Family Relationships/Functioning
Social Participation/Support
Free-text Observations

Sample Output Description
Output Type:

Structured, actionable case summary and domain assessments—automatically generated.

Output formats: Excel report, consolidated CSV, audit-ready PDF, or summary email.

Sample Output Example:

Consolidated Report Sections:

Strengths (by domain): e.g., “Stable employment (Income Security); Good school attendance (Education)”
Risks (by domain): e.g., “Chronic health condition (Health); Overcrowded housing (Housing)”
Goals: e.g., “Secure better housing; Access medical support; Improve school participation”
Tasks: e.g., “Refer to housing services; Schedule doctor appointment; Arrange tutoring support”
Domain Tagging: Each item clearly labeled with its domain
Unique Dates: All unique ‘Engagement Dates’ consolidated for reference

Summary Paragraph: High-level overview of the family’s situation and needs

Family Support Score: Automated rating (e.g., 65%) with interpretation band (“Moderate Need”)
Special Notes: Any urgent risks or HITL flags
Final Email or Dashboard Entry:Contains summary, score, and recommendations for social worker review and approval.

Other resources

Presentation: WelfareWise.ai.pptx - Google Slides

Agentic Orchestration Maestro Flow: https://drive.google.com/file/d/1IsK-UeeFMwKeMzYfsWJYgHibQnosBcts/view?usp=sharing

6 Likes