Inferenz helps providers, payers, MedTech and life sciences organisations move from reactive operations to intelligent and value-based care leveraging data and AI.
Seeking to streamline the assessment workflow, the organization needed an efficient, scalable solution that converted caregiver-patient phone assessments directly into structured, actionable records.
Call-Based Assessments Caregivers performed patient evaluations via phone, manually entering details into PDFs
Manual Data Entry Challenges Filling PDFs by hand created bottlenecks, increased risk of errors, and delayed care reporting
Lost Analytics Value Data trapped in static PDFs made it difficult to analyze patient outcomes and performance trends
Operational Inefficiency Lack of real-time, structured data restricted rapid decision-making and compliance monitoring
Our Solution
Inferenz implemented an AI-powered system that automatically transcribes assessment calls, generates filled digital PDFs, and extracts all key information into database-ready formats supporting analytics and reporting.
Discovery & Assessment Mapping Workshops with client teams identified relevant assessment questions and standardized field definitions
Dynamic PDF Generation System instantly mapped transcribed answers to the correct PDF fields, creating fully filled digital forms with zero manual input
Integrated Quality Validation Automated checks ensured completeness, consistency, and compliance of all filled forms and corresponding datasets
Cloud-Based Workflow Entire solution deployed on scalable AWS infrastructure, integrating with downstream reporting and EMR systems
Automated Call Transcription AI models transcribed caregiver-patient phone conversations, distinguishing speakers for accurate data mapping
Structured Data Extraction All assessment data extracted from calls and PDFs stored in a normalized database format for real-time analytics
Review and Edit Step Caregivers could review and update AI-generated PDFs before submission, ensuring clinical accuracy
The client held millions of trial, lab, and EHR records but no single view to run predictive risk analytics. The gap slowed their AI in healthcare push and let early warning signs slip past care teams. It faced certain issues like:
Fragmented Inputs Vital signs, lab results, and trial data lived in separate siloes making doctors ignorant about cross-source trends in time
Fuzzy Definitions “Positive” patient labels varied by unit and study indicated lack of clear rules causing noisy training targets
Class Imbalance Only ~4 % of records showed early signs of decline featuring standard models that over-fit to the safe majority
Our Solution
We built a cloud data hub that pulls trial, lab, and EHR records into one stream, ready for agentic AI agents to scan in near real time. We also trained a risk model that fuels an early alert system, sending care teams clear flags up to six months before trouble hits
Data preparation We standardized units, aligned IDs, and filled important gaps to give the model clean, consistent inputs.
Balancing rare events We adjusted training samples so early-risk cases stayed visible, improving the model’s sensitivity.
Model development An iteratively tuned ensemble found patterns that generalize well to new patients.
Validation Cross-validation plus a reserved test set confirmed strength before rollout.
Handling missing data Sound statistical methods replaced blanks, keeping key signals intact while avoiding bias.
Feature engineering Rate-of-change, rolling average, and trend indicators captured subtle shifts in patient status.
Interpretability layer Clear factor-importance views help clinicians see why each alert fires.
The client, a US-based mentor network, runs nationwide programs that pair mentors with people who need daily support. Yet critical data sat in two siloes: a legacy SQL store and Salesforce. The split created three urgent gaps
Data Integration No common keys tie Salesforce and SQL together, which resulted in duplicate or missing mentor records leading the staff to hunt through both databases and risk overlooking history
Reporting A sluggish export-import routine pulled Salesforce data into SQL once a day driving manual effort that forced teams to lose hours each week building simple metrics
Data Quality Names, addresses, and IDs differ between sources as leaders questioned the numbers, slowing decisions and raising red flags with compliance reviewers for data risks
Our Solution
Central Master Repository A new SQL Server hub now stores the golden record for every mentor
Data Vault Layer Hub, Link, and Satellite tables keep history while preserving source-level detail
Hash-Key Generation Unique keys for each record end merge conflicts and speed look-ups
ETL with SSIS Incremental jobs load, cleanse, and validate 1 M rows on a defined schedule
Self-Service Analytics Power BI dashboards surface up-to-date mentor counts, placements, and gaps without manual extracts
The client offering health and nutrition products globally realized that its legacy DB2 system could no longer support growing business demands. Data was scattered across several systems leading to following challenges in:
ERP No unified view of products, SKUs, customers, or sales channels, limiting strategic insights
Marketing Inability to measure campaign effectiveness and allocate spend efficiently, leading to poor conversions.
Pricing systems Lack of real-time updates created a rigid, outdated pricing structure.
Inventory management Frequent stockouts and ageing inventory due to missing expiry tracking.
Our Solution
We adopted a phased, consultative approach to modernize the client’s data landscape:
Centralized Data Warehouse Built a unified warehouse with integrated facts (orders, shipping) and dimensions (product, customer, region, channel) aligned to SKU and product hierarchies.
Forecasting & Inventory Optimization Deployed predictive models to reduce stockouts, manage SKU expiry, and drive efficient pricing strategies.
Data Unification Integrated siloed sources—ERP (Mozart), pricing systems, marketing data (Adobe, Google, Yahoo, Amplitude, Creteo), and Revionics—into a single source of truth.
SQL to Snowflake Migration Transitioned legacy SQL systems to Snowflake for scalable, cloud-native architecture.
BI Modernization Upgraded from Sisense to Power BI after evaluating performance and usability against Tableau.
Customer Intelligence with Marketlo™ Used Inferenz’s proprietary platform to segment and predict customer behavior (e.g., 80% aged 60+), enabling 0–100 centile-based targeting and optimized catalogue delivery.
The client pursued rapid growth through strategic mergers and acquisitions (M&A), acquiring 12 homecare entities. Each acquisition brought its own distinct payroll, EMR, and HR systems. This led to several challenges like
No Single Source of Truth across payroll, EMR, and HR systems
Siloed, inaccessible data, delaying decision-making and ROI
Inconsistent KPI tracking across 30+ metrics for clinical, scheduling, and compliance
Our Solution
We developed a tailored M&A data integration framework focused on speed, scalability, and data integrity; comprising of
Discovery Workshops Aligned with client teams to map key metrics and use standardized definitions across systems
Snowflake-Powered Architecture Enabled fast, scalable onboarding of new acquisitions
AI-Powered De-Duplication Framework Ensured accurate ROI tracking by distinguishing overlapping client records
Automated Data Quality Checks Delivered clean, consistent, and compliant datasets using data quality checks
