NLP for Superior User Interaction

Client at a glance

$130M+Raised in venture capital
3,000+Customers globally
$10B+Managed outcomes

Advancing User Experience with AI-Powered Insights

Quantive (formerly Gtmhub) is a leading provider of strategy execution software and services, built on the Objectives and Key Results (OKRs) methodology. Their platform helps organizations achieve alignment, enhance visibility, and foster a results-driven culture. Quantive’s platform now offers an enhanced user experience powered by real-time semantic suggestions and predictive analytics. By combining cutting-edge NLP technologies with scalable infrastructure, we empowered Quantive to deliver smarter, faster, and more intuitive interactions that drive customer success.

Challenge

As digitalization drives innovation at unprecedented speeds, Quantive faced the challenge of optimizing its product to offer personalized and seamless user experiences. Specifically, they sought to:

  • Deliver customized recommendations for OKR creation and management, alleviating the complexity of self-service workflows.
  • Streamline navigation to Insights, ensuring users could quickly access the most relevant information.
  • Boost customer satisfaction by automating form fill-ins and improving usability.

Our Approach

To address these needs, we collaborated with Quantive to deliver a data-driven solution that leverages advanced machine learning and natural language processing (NLP) technologies.

Key actions included:

  1. Data Infrastructure:
    • Developed a PostgreSQL database connected to an automated pipeline in Azure Cloud.
    • Integrated Azure Data Lake and Data Factory for sourcing data from MongoDB, later orchestrated through Azure Synapse for scalability.
  2. NLP Integration:
    • Built an LSTM-based neural network using TensorFlow for automatic form fill-ins.
    • Deployed a pre-trained Transformer model to generate real-time semantic recommendations for Insights, integrated via gRPC microservices fetching data from a Kafka stream.
  3. Operationalization of ML Pipelines:
    • Automated data science workflows using MLflow, Jenkins, and SonarQube for model training, testing, and deployment.
    • Set up Grafana dashboards to monitor real-time service accuracy and performance, ensuring ongoing reliability.

Impact Delivered

  • Precision in Automation: Achieved over 95% accuracy in automatic entity recognition and form completion.
  • Enhanced Engagement: Significantly boosted customer engagement by providing timely, relevant recommendations.
  • Increased Usability: Improved adoption of OKRs and Insights features by streamlining workflows and enhancing user satisfaction.
  • Performance Excellence: Delivered a model service with a mean response time of 150 milliseconds per request, seamlessly handling thousands of customer interactions daily.

Expertise and Scope

  • Deliverables: NLP-based recommendation system, automated data pipelines, monitoring dashboards
  • Technology Stack: Python, TensorFlow, Keras, MLflow, Docker, Jenkins, SonarQube, Azure Data Lake, Azure Data Factory, Azure Synapse, PostgreSQL

Preparing a Tech Platform to Scale Goodlord Operations

Client at a Glance

£2B transactionsprocessed annually
65K+ tenants and landlordsusing the platform
50%+YoY growth in platform usage

Empowering Goodlord to Scale with Confidence

Goodlord is a UK-based software service provider offering innovative solutions for landlords, agents, and tenants. The company specializes in simplifying and digitizing the rental process, including:

  • Preparation and maintenance of rental documentation, such as electronic forms and contracts.
  • Facilitating utility services like gas, electricity, and water connections.
  • Streamlining workflows with electronic signatures, invoices, and automated notifications for contract extensions and terminations.

Goodlord’s platform supports seamless online payments and provides a fully digital experience for all parties involved in the rental process.

Through a modernized architecture, enhanced processes, and seamless integration of remote teams, Goodlord is now equipped to handle rapid growth and deliver exceptional service experiences. By enabling scalability and adaptability, we helped Goodlord strengthen its leadership in the property management software market.

Challenges

Goodlord faced a critical need to modernize its web-based platform by:

  • Migrating legacy functional PHP code to a more scalable and maintainable architecture using Symfony.
  • Implementing new features, improving performance, and integrating tests to ensure quality.
  • Collaborating with external remote engineers for the first time—an ambitious step for a company with an established onsite-only work culture in London.

Goodlord sought experienced software developers who could not only integrate into their in-house processes but also propose innovative solutions to address support issues and optimize the platform. This was a high-stakes endeavor, requiring seamless collaboration and effective communication between distributed teams.

Our Approach

To help Goodlord achieve its goals, we began with an in-depth analysis of the platform’s existing architecture. By engaging with onsite engineers, support teams, and product managers, we gained a comprehensive understanding of the business objectives and technical challenges.

Key actions included:

  • Modernizing the Architecture: Migrated legacy functional PHP code to object-oriented programming (OOP) modules using Symfony, enabling greater scalability, maintainability, and faster implementation of new features.
  • Optimizing the Database: Reorganized the MySQL database structure to enhance performance and efficiency.
  • Improving the User Interface: Transitioned the platform’s web interface to React components, delivering a faster and more intuitive user experience.
  • Enhancing Quality Assurance: Integrated comprehensive unit and end-to-end tests to simulate various loads, prevent defects, and ensure stability.
  • Data Integration Support: Developed ETL processes using SSIS to seamlessly import end-customer data from various systems into Goodlord’s platform.

Impact Delivered

  • Scalable Architecture: Introduced a Symfony- and Scala-based architecture that supports future platform growth.
  • Improved Database Performance: Optimized the MySQL database for enhanced reliability and speed.
  • Elevated User Experience: Delivered a modern web interface with React components, improving performance and engagement.
  • Robust Quality Assurance: Achieved comprehensive test coverage, ensuring defect prevention and stability.
  • Successful Remote Integration: Enabled effective collaboration with external remote engineers, establishing a new operational model for Goodlord.

Expertise and Scope

  • Technology Stack: Symfony, Scala, React, MySQL, Concourse, Travis, AWS
  • Focus Areas: Software migration, multi-tech collaboration, performance optimization, data integration

Unlocking Customer Insights for Kaufland

Client at a glance

1530Locations globally
40 yearsFounded in 1984
€34.2BRevenue in 2023

Redefining Retail Excellence with Data-Driven Insights

Operating over 1,500 stores and warehouses across eight countries, Kaufland is a leader in the FMCG sector, with an online marketplace in Germany complementing its expansive physical presence. By leveraging advanced analytics and predictive modeling, Kaufland has transformed its approach to customer engagement, ensuring it remains a trusted choice in the highly competitive retail market.

Challenge

Kaufland, a leading international retailer in the fast-moving consumer goods (FMCG) sector, sought to gain deeper insights into their customers’ preferences, habits, and sentiments to optimize satisfaction and deliver tailored products and services. Without a comprehensive understanding of their customer base, they risked falling short of expectations, leading to dissatisfaction and inefficient use of resources. To stay ahead in a highly competitive market, Kaufland needed a solution to unlock actionable customer insights and improve their strategic decision-making.

Our Approach

To address Kaufland’s challenges, we deployed advanced analytics and customer segmentation techniques, enabling a more granular understanding of their customer base and market dynamics.

Key actions included:

  • Customer Segmentation: Utilized methods such as K-means clustering, DBSCAN, Regression Analysis, ANOVA, and PCA to segment the customer base by attributes and preferences, revealing distinct behavioral patterns.
  • Predictive Modeling: Developed predictive models using XGBoost and Random Forests to forecast customer behavior, recommend personalized actions, and optimize marketing strategies.
  • Loyalty Programs: Designed tailored loyalty programs to enhance customer retention and engagement, driven by insights into specific customer segments.
  • KPI Tracking: Established robust systems for monitoring and analyzing key performance indicators, ensuring transparency and data-driven decision-making.

Impact Delivered

  • Enhanced Customer Understanding: Delivered actionable insights into customer segments and behaviors, enabling more targeted and effective marketing campaigns.
  • Improved Retention: Tailored loyalty programs significantly boosted customer engagement and retention rates.
  • Data-Driven Decisions: Built a foundation for strategic decision-making through transparent KPI tracking and performance evaluation.
  • Optimized Resources: Enabled more efficient allocation of resources by aligning products and services with customer needs.