TMLink connects disparate data records across systems using advanced machine learning — so you can gain a complete, accurate view of your customers, patients, and citizens.
In today's data-driven world, organizations face a significant challenge: data fragmentation. With information scattered across multiple sources, systems, and formats, it's difficult to get a complete and accurate view of the data.
TMLink helps overcome this challenge by connecting the dots — enabling businesses and institutions to uncover new insights, improve decision-making, and drive innovation.
A streamlined, three-stage pipeline that turns fragmented datasets into connected intelligence.
Securely log in and upload your record linkage dataset. TMLink supports various entity information including names, birthdates, and IDs. The platform's frontend, powered by Streamlit, uses APIs to manage backend logic and data storage.
Select the columns needed for linkage and deploy TMLink's advanced probabilistic models fused with transformer embeddings. Records are matched in minutes based on names, addresses, dates of birth, and more — with processing time proportional to dataset size.
Use basic or advanced search to query your linked data. TMLink generates ad hoc queries based on probabilistic linkage thresholds and supports Text-to-SQL for advanced users. Spelling variations and minor discrepancies don't hinder the linkage process.
Connecting data unlocks insights that improve every corner of your organization.
Eliminate duplicates, correct errors, and enhance overall data accuracy by linking records across sources into a single, authoritative view.
Gain a comprehensive 360° view of customer behavior, preferences, and needs by linking records across all your data systems.
With accurate and complete data, organizations can make informed decisions and drive business growth backed by reliable intelligence.
Meet regulatory requirements and manage risk by identifying potential data breaches, security threats, and compliance gaps through linked records.
TMLink powers connected data strategies across sectors where accuracy and completeness are mission-critical.
Link patient records across hospitals, clinics, and labs to improve care coordination and patient outcomes — even when records use different formats or contain data entry errors.
Connect customer data across accounts, transactions, and institutions to enhance risk management, fraud detection, and regulatory compliance.
Link citizen data across departments and agencies to improve public services, policy-making, and program delivery at scale.
Understand customer behavior and preferences across channels to drive targeted campaigns, personalization, and business growth.
A concise guide to the current TMLink runtime, application stages, merge behavior, and operating commands.
This documentation covers the application flow from CSV upload through record linkage and entity search, along with the current container-based startup model.
TMLink runs as a Streamlit frontend with a FastAPI backend inside Docker. The product flow stays intentionally simple:
Accepted uploads are merged before linkage runs, so the workflow operates on the current combined dataset rather than only the last file added.
Use master_setup.sh as the main entry point. It rebuilds the application packages, builds the Docker image, starts the container, and waits for the services to come up.
chmod +x master_setup.sh
bash master_setup.shhttp://localhost:8501 serves the Streamlit frontend.http://localhost:8000 serves the FastAPI backend.Users upload one or more CSV files. Compatible files are merged into the current working dataset and the summary updates immediately.
The Start TMLink action runs against the (merged) Stage 1 dataset. Users must choose between 5 and 10 linkage fields.
After linkage completes, users can search linked records using basic search or advanced query generation backed by the saved Stage 2 similarity output.
# Build image
docker build -t taiwotman/tmlink:latest .
# Run container
docker run -d -p 8501:8501 taiwotman/tmlink
Port 8501 already allocated
This means an older container is still bound to the frontend port. Stop and remove the old container before starting a new one.
Discover how TMLink can help your organization unlock the power of connected data. Schedule a personalized walkthrough with our team.
Fill in the form and we'll get back to you within 24 hours to confirm a time that works for you.
By submitting, you agree to be contacted by TMLink / BigCodeGen. No spam, ever.