CVM Stories podcast on SpotifyCVM Stories podcast on Apple

Let us introduce Justas Jankūnas, a Team Lead at Exacaster, who has a unique blend of business and technical know-how. Justas shares his insights and strategies for the world of data analytics, Machine Learning (ML), and Customer Value Management. He discusses the crucial collaboration between analytics and CVM teams, the impact of ML in telecom, and the challenges and best practices in implementing ML operations (MLOps). Justas draws inspiration from the world of Formula 1 racing, where precision, strategy, and teamwork prevail. Tune in as Justas Jankūnas takes us backstage and lays out his predictions for the future of ML in CVM and beyond.

10 Key Takeaways:

  1. Teamwork in Data Analytics: Tight cooperation between data analytics and CVM teams is like a Formula 1 racing team. Just like a successful Formula 1 team requires coordination among drivers, engineers, and strategists, effective teamwork is crucial for the success of data analytics projects in telecom. It’s not just about individual skills; it’s about how well the team works together.
  2. Role of Analytics Engineer: The role of an Analytics Engineer blends business analysis with data engineering skills, bridging the gap between technical and business aspects.
  3. Iterative Collaboration: Successful collaboration between analytics and CVM teams happens through iterative cooperation, with technical experts communicating in understandable language.
  4. Machine Learning in Telecom: ML plays a crucial role in predicting customer behavior, optimizing network operations, and enhancing customer support through chatbots and fraud detection.
  5. Customer Behavior Prediction: ML models are used to predict customer behavior, such as prepaid to postpaid migration, churn, upselling, and cross-selling, yielding significant business results.
  6. Balancing Tech with Business Goals: Balancing technical expertise with a deep understanding of business dynamics is similar to a Formula 1 driver understanding both the mechanics of the car and the race strategy. Just as a driver must communicate with engineers to optimize the car’s performance, technical experts in data analytics must communicate effectively with business teams to align projects with company objectives.
  7. Data Quality Challenges: Data quality is a common challenge in ML projects, and early steps like data quality assessment and exploratory data analysis are crucial.
  8. Clear Communication: Communicating complex ML concepts in simple terms is essential, Justas emphasizing the use of everyday analogies.
  9. Machine Learning Operations (MLOps): MLOps is critical for deploying ML models effectively in production, ensuring stability, monitoring, and efficient model updates.
  10. Domain-Specific Challenges: Different domains in telecom, like credit scoring, mobile upsell, churn, and cross-selling, require careful definition of target variables and understanding specific challenges.

Reading Recommendations:

Explore these insights to gain a deeper understanding of data analytics, ML, and their applications in the telecom industry. Justas’ experience and approach provide valuable guidance for professionals in the field.