Pravin Wadekar

Experienced Technology Leader and Innovator with a proven track record of driving innovation, delivering scalable technology solutions, and building high-performing teams. Specialized in developing industry-specific cloud platforms, no-code solutions, and machine learning applications for sectors including automotive, IoT, and retail. Known for optimizing operations, accelerating revenue growth, and enabling digital transformation.

• Secured a $1.5M commitment for a new platform at Excellon Software.
• Designed a no-code workflow engine, cutting change request time by 30% at Excellon Software.
• Standardized the Data model and Migrated legacy systems to microservices, reducing operational costs by 30% at Excellon Software.
• Orchestrated a cultural transformation within the company, shifting the focus from application-centric to platform-centric operations, fostering innovation and scalability across all technology initiatives at Excellon Software.
• Owned and Executed R&D effort for new platform DMS platform with integrated tooling to reduce the support cost substantially.
• Raised $1M in seed capital for Smartalyse Technologies' IoT Gateway.
• Achieved 18% YoY growth at Microsoft India, driving $21M in revenue.
• Built Onsight-offshore business of 6+ M and 130+ strong team at Persistent Systems.

Building a product from an initial concept to full deployment is a journey filled with challenges, opportunities, and learnings. For me, this journey began with a deep realization: infrastructure costs were unsustainable, customer demand for scalability was increasing, onboarding time needed significant improvement, and the nature of our business demanded a long-term, strategic approach. Additionally, transforming a brownfield business and reshaping organizational culture were critical to ensuring success. This is the story of how these factors shaped my approach to building a scalable and future-proof product.

Like many companies in the enterprise space, we faced rising infrastructure costs. Our existing setup was not optimized for scalability, and every incremental growth in customers added disproportionate overhead. Customers were demanding faster onboarding and seamless scaling. The realization hit: we needed a new approach.

Key Questions We Asked:

• How can we reduce infrastructure costs without sacrificing performance?
• What architecture allows for true horizontal scalability?
• How do we accelerate customer onboarding?
• Can we design a solution that grows with evolving business needs?

Extensive market research and competitive analysis revealed that business processes and priorities change continuously. Understanding competitor strategies and customer pain points became instrumental in refining our approach.

Findings from Market Research:

• Flexibility is Key: Enterprise customers need solutions that can adapt quickly to evolving needs.
• Friction in Onboarding: Many competitors struggle with prolonged onboarding times, creating a barrier to adoption.
• Data-Driven Decision Making: Companies leveraging AI and real-time analytics outperform those relying on traditional reporting.
• Customer Support & Transparency: A significant gap exists in providing clear, real-time insights into product performance and cost structures.

This research validated our need to prioritize adaptability, seamless onboarding, and observability while designing our solution.

Recognizing the need for faster turnaround times and improved issue resolution, we pivoted to build a no-code platform. This approach enables customers to configure and deploy solutions with minimal technical expertise, reducing dependency on engineering teams and expediting time-to-value.

Benefits of the No-Code Approach:

• Rapid Deployment: Reduces onboarding time and accelerates time-to-market.
• Empowering Business Users: Allows non-technical users to configure workflows without engineering intervention.
• Scalability & Flexibility: Adapts to evolving business processes with minimal friction.
• Improved Time to Resolution: Enables faster fixes by automating issue identification and response workflows.

By shifting towards a no-code paradigm, we ensured that enterprises could build, modify, and scale solutions without being slowed by traditional development cycles.

As the product evolved, we changed our approach by integrating user stories for each persona who would be using the application. This strategy significantly helped in consolidating functionality and led to the creation of dedicated applications for different user groups:

• Customer App: Focused on enhancing customer interactions, streamlining support, and delivering a seamless user experience.
• Business Users App: Tailored for operational efficiency, data-driven insights, and managing complex workflows.

This shift improved the usability and adoption of the platform while optimizing business workflows, API architecture, and security considerations for these applications, ensuring robust and seamless interactions across all integrated systems.

A customer-centric approach is critical while building any platform or solution. Without prioritizing customer feedback, real insights into usability, adoption barriers, and feature effectiveness remain elusive. We involved customers at every step of our journey to gather feedback, conduct field trials, validate solutions, and optimize them based on real-world use cases. By embedding continuous feedback loops and fostering open dialogue with users, we refined features, enhanced user experiences, and drove overall product success.

Customer demand for scale varies based on multiple factors, including:

• B2B2C Network Size: The extent of their business relationships impacts scalability needs.
• Retail and Distribution Network: The number of products in circulation affects required infrastructure.
• Seasonal Demand: The ability to withstand peak demand periods is crucial.
• Customer Size & Commitment: Pricing structures and baseline commitments influence how businesses scale within the network.

These factors led us to design a workflow fabric architecture that supports dynamic scaling. Additionally, we introduced a no-code API design studio, reinforcing an API-first approach. This ensures:

• Seamless extensibility for varying customer needs.
• Faster integration with existing ecosystems.
• Optimized workflows for businesses with fluctuating demand.

The next stage in our product journey involves moving the workflow fabric to AI-driven agents within a no-code development studio. This transition will:

• Enable more intelligent and automated workflow execution.
• Allow users to configure AI-driven process automation seamlessly.
• Reduce operational inefficiencies by predicting and optimizing business processes.

Additionally, we are building a process designer studio to enhance workflow creation flexibility and efficiency. We will also release our application design studio to enable rapid prototyping and deployment of customer and business applications.

To expand the platform's capabilities, we are launching a connector factory platform that will facilitate seamless integrations with third-party services, enterprise systems, and ecosystem partners. Driving this initiative forward requires an aggressive focus on partnerships, ensuring that our platform is well-integrated into the broader enterprise software landscape.

Building a product is more than just engineering—it’s about aligning vision, strategy, and execution. For anyone embarking on a similar journey, my advice is simple: stay agile, listen to your customers, and embrace the challenges that come with scaling a business. The rewards of seeing an idea materialize into a successful product are well worth the effort.

Are you facing similar challenges in your product journey? Let’s connect and exchange insights!

Patent : SYSTEM AND METHOD FOR CONFIGURING IOT DEVICES | https://patents.justia.com/inventor/pravin-wadekar

Introduction

Building a robust platform architecture to support large-scale, data-intensive applications requires careful planning, meticulous design, and a deep understanding of both current and emerging technologies. This blog explores the technical decisions that helped create a scalable, no-code, microservices-based platform, incorporating advanced observability, real-time analytics, and a flexible connector ecosystem. By leveraging cutting-edge tools and proven architectural patterns, this platform not only streamlined operational efficiency but also delivered business-critical insights and enhanced customer experiences.

Microservices and No-Code: A Dual-Pronged Approach

At the core of the architecture lies a distributed microservices framework designed for horizontal scalability and fault isolation. Each service was independently containerized using Kubernetes, enabling rolling updates, rapid scaling of individual components, and seamless failover handling. To ensure high availability, we implemented a multi-region deployment strategy, leveraging managed services like Amazon EKS, backed by autoscaling groups and load balancers to handle unpredictable traffic spikes.

The no-code layer introduced a business-friendly abstraction over the microservices. It was built using a custom low-code engine underpinned by a rule-based orchestration framework. This allowed non-technical stakeholders to define workflows, event triggers, and decision trees, all of which compiled down to optimized microservices calls. The result was not just reduced implementation time, but also a significant decrease in maintenance overhead since workflows could be updated on-the-fly without redeploying any core services.

Observability: A Proactive Support Paradigm

Traditional observability tools often focus on application performance metrics, but we went beyond the standard telemetry stack. By integrating Prometheus for real-time metrics aggregation, Grafana for dynamic visualization, and distributed tracing with Jaeger, we created a system where every request could be traced across microservices. This setup provided detailed visibility into latency hotspots, service-to-service dependencies, and contention points.

More importantly, we leveraged this observability framework to build predictive analytics for the support team. Using a combination of machine learning models deployed on AWS SageMaker and data pipelines orchestrated by Apache Airflow, we could forecast potential system bottlenecks and predict user-facing issues. Alerts were sent to our incident response system (integrated with PagerDuty), enabling support engineers to preemptively address problems before they impacted end users.

Building a No-Code Data Pipeline

To support real-time and batch data workflows, we constructed a no-code data pipeline using a Lakehouse architecture. Data was ingested through Apache Kafka, ensuring scalable and fault-tolerant streaming. Once ingested, data was stored in MinIO, a high-performance object store compatible with the S3 API. Metadata management and schema evolution were handled by Apache Hive, which provided a flexible interface for querying and cataloging data.

For distributed query execution, we integrated Trino, enabling SQL-based analytics on top of the Lakehouse. This approach allowed us to maintain a single source of truth while supporting both ad-hoc queries and pre-defined transformations. To enhance the no-code experience, we developed a pipeline builder UI that leveraged these technologies under the hood, allowing users to drag-and-drop components, configure streaming transforms, and deploy workflows without writing code. The result was a scalable ETL solution capable of handling everything from continuous streams to high-volume batch jobs.

Driving Business-Specific Insights

Data pipelines became the backbone for business-specific machine learning and AI models. Models were trained on historical data stored in the Lakehouse and deployed as inference endpoints using Kubernetes-based model servers. Examples included:

• Demand Forecasting: Using historical sales data and external market signals, we built LSTM-based models to predict inventory requirements and ensure optimal stock levels.
• Customer Footfall Prediction: Computer vision and sensor data were integrated into the pipeline, providing real-time estimates of customer traffic at retail locations.
• Lead Scoring: A gradient boosting model, fine-tuned with LightGBM, was implemented to rank sales leads by quality, enabling more targeted marketing campaigns.
• Real-Time Analytics: Using Apache Pinot, we provided sub-second query capabilities on streaming data, delivering live dashboards that updated as events occurred.

By leveraging these data-driven insights, the platform transformed into a decision-making engine, enabling customers to optimize operations, reduce waste, and increase profitability.

Connector Ecosystem and Seamless Integration

A modern platform cannot exist in isolation. We designed a connector factory framework to simplify integration with a wide range of external systems. Using a modular adapter pattern, connectors were built for:

• Enterprise Resource Planning (ERP) Systems: Adapters for SAP and Oracle allowed real-time synchronization of inventory, production schedules, and financial data.
• Insurance and Loan Providers: Secure API integrations ensured that service orders could be validated and underwritten in real time.
• Payment Gateways: Integration with Stripe, PayPal, and regional payment providers enabled seamless transactions across multiple currencies.
• Communication Channels: Connectors for email (via SendGrid), messaging (via Twilio), and collaboration platforms (such as Slack and Microsoft Teams) ensured that notifications, alerts, and customer communications flowed smoothly.

The connector factory was augmented by a built-in schema registry and a data migration framework. This made onboarding new customers straightforward by automating schema alignment, data transformation, and initial data loading processes. Combined with continuous integration and continuous delivery (CI/CD) pipelines, the entire ecosystem could evolve without downtime, ensuring high availability and reliability.

Conclusion

Building a scalable, high-performance platform architecture requires more than just picking the right tools—it demands careful orchestration of technologies, robust design principles, and a forward-thinking approach. By combining microservices with a no-code framework, integrating advanced observability, leveraging a Lakehouse architecture for flexible data pipelines, and creating a versatile connector ecosystem, we achieved a platform that meets complex business needs, reduces operational costs, and continuously improves customer satisfaction. This architecture serves as a foundation for future growth, ensuring that it can evolve alongside emerging trends and increasing demands.