Soumyajit Sarkar
AI Solutions Architect and Backend Analytics Engineer
Twenty years building analytics and AI systems that run at scale. One of them has tracked more than $100 million in revenue across 80 million users a year.
Table 1. Scale of work, counted
20+1 years of experience
= 1 year
80M+2 users analyzed annually
= 2 million users
$100M+2 revenue tracked
= $5 million
7-82 years of continuous analytics
= 1 year; the cut mark shows the range
- 1 Career total as published by the author.
- 2 Backend analytics management system architected and maintained by the author; figures as published by the author.
Table 2. Selected systems and outcomes
Each row is a system he designed or built. Results are as published by the author.
| System | What it does | Result |
|---|---|---|
| Backend analytics management system | Real-time data processing, automated reporting and predictive analytics for business intelligence on high-traffic applications. | $100M+ revenue tracked 80M+ users analyzed annually |
| Enterprise AI solutions architecture | ML pipelines, real-time inference systems and automated model deployment frameworks for multiple Fortune 500 companies. | Reduced infrastructure costs by 40% Improved model performance by 65% |
| EduCLI.com educational analytics | AI-driven learning insights, personalized content recommendations, learning-path optimization and assessment analytics. | 85% improvement in student engagement 92% learning path optimization success 78% increase in course completion |
| AI-enhanced luxury rental platform | Recommendation engine, predictive pricing, AI pricing optimization and automated customer-behavior analysis with real-time market analysis. | No figure published |
Table 3. Specialisms
Three areas of practice, each with its own detailed table.
-
3a AI Solutions Architecture
Enterprise AI system design, ML pipeline architecture, real-time AI systems and AI infrastructure optimization.
-
3b Backend Analytics Engineering
Real-time data processing, revenue tracking systems, data pipelines, analytics APIs and data warehouse design.
-
3c Educational Analytics Platforms
Learning analytics, student performance tracking, learning-path optimization and performance prediction.
Table 4. Tooling register
- AI and machine learning
- TensorFlow, PyTorch, Scikit-learn, NLP, computer vision and deep learning, neural networks, MLOps
- Backend analytics
- Python, Node.js, Go, Java, Scala; data pipeline architecture, real-time analytics, ETL/ELT
- Data engineering
- Apache Kafka, Apache Spark, Snowflake, BigQuery, stream processing, data modeling
- Databases
- PostgreSQL, MongoDB, Redis, Elasticsearch, ClickHouse
- Educational analytics
- Learning management systems, student performance analytics, adaptive learning algorithms, educational data mining
- Cloud and DevOps
- AWS, Google Cloud, Azure, Docker, Kubernetes, CI/CD, infrastructure as code
- ML operations
- MLflow, Kubeflow, Apache Airflow, Prometheus, Grafana, model deployment and monitoring