Sujal Anand

Integration Architect and Developer with 6+ years of experience delivering scalable, secure, and high-performance integration solutions using MuleSoft and Salesforce. Expert in API-led connectivity, microservices, and cloud-native deployments (CloudHub, RTF, AWS). Proven track record in leading end-to-end SDLC, collaborating with cross-functional teams, and mentoring developers. Passionate about leveraging emerging technologies like neural search, RAG, and AI-driven solutions to drive business impact.

Database ↔ SFMC Integration (ETL | File-based | Incremental Loads) : Qualtrics Survey Integration (Event-Driven | Micro-service | Metadata-Driven Mapping)

• Designed, developed, and deployed an end-to-end MuleSoft solution to extract comprehensive customer data—including Linode customer information, services availed, and support tickets raised—from multiple database tables using SQL joins, efficiently handling a historical load of 70 million records as well as subsequent incremental data loads to ensure up-to-date information.
• Transformed and securely pushed the extracted data files to SFMC provided SFTP.
• Developed and scheduled Automation Studio jobs within SFMC to automate pulling multiple data files from SFTP and ingesting them into multiple Data Extensions, managing both the initial historical load and subsequent incremental loads running as soon as files are dropped to SFTP.
• Architected the integration to be highly configuration-driven, offering flexibility such as the ability to disable support ticket synchronization without affecting other integrations like customer or services sync. Furthermore, new datasets from the database to SFMC Data Extensions can be added solely through configuration changes, eliminating the need for MuleSoft code modifications. Even scheduling parameters like the Mule Cron expression can be adjusted externally. This approach significantly accelerated release cycles and minimized reliance on development resources.
• Implemented comprehensive monitoring and alerting using Splunk, enabling proactive issue detection and resolution. Created detailed runbook and onboarded the operations team to ensure smooth production support and incident handling.
• Designed and Developed a second integration phase to extract journey engagement metrics from SFMC—including emails sent, clicks, views, unsubscribes, journey versions..etc—incrementally exporting these metrics to SFTP. Subsequently, MuleSoft jobs pull these files, process them, and ingest the data into the Oracle Enterprise Data Warehouse for downstream reporting and analytics.
• Each integration is linked to a dedicated state table in the Oracle that captures detailed metadata, including integration name, total records processed, success and failure counts, error messages, start and end timestamps, and overall integration status (success/failure).

Database → Salesforce, Marketo, and Lytics Integration (Incremental Load):

• Designed and developed an incremental data integration solution to extract customer data from the database and determine customer status using complex SQL-based winning logic.
• Synchronized customer IDs and calculated status to Salesforce leveraging Salesforce Bulk API v2 for efficient large-scale data processing.
• Implemented automated alerts to notify operations teams of failed records for prompt resolution.
• Developed robust error handling and retry mechanisms to reprocess and resync failed records, ensuring data integrity.
• Enabled downstream synchronization of customer IDs from Salesforce to Marketo and Lytics using out-of-the-box (OOTB) sync capabilities, maintaining consistent customer data across marketing platforms.

• Designed and developed an event-driven integration to process survey data from Qualtrics Survey API, transform it, and publish to a JMS topic for parallel processing.
• Built a fan-out architecture where one consumer pushes data into a relational database, and another routes it to Salesforce for operational workflows and reporting.
• Externalized transformation logic using Salesforce metadata, enabling non-code updates to mapping rules and reducing release cycles.
• Designed and implemented Salesforce automation using Apex Triggers to detect low survey ratings and automatically notify the respective support engineers via email.

Marketo Integration (Delta Sync)

• Led the project end-to-end, including requirements gathering, design, development, testing, and deployment; collaborated with stakeholders to align integration goals.
• Onboarded and mentored a junior developer, providing guidance on MuleSoft development, integration best practices, and project standards.
• Designed and implemented a daily incremental data load process from Enterprise Data Warehouse (EDW) to Marketo using REST APIs.
• Built logic to extract only delta records based on timestamps, transformed the data as per Marketo schema, and ensured reliable data ingestion via Marketo APIs.
• Designed the integration to be configuration-driven, allowing seamless handling of frequent field additions and query changes without code modifications—enabling rapid delivery of enhancements within a week instead of a typical monthly release cycle.

6sense Integration (Large File Ingestion – ~4 GB per File):

• Designed, developed, and deployed a file-based integration to process large datasets (~4 GB per file) from 16 inbound files on an SFTP server and ingest them into a relational database.
• Implemented a robust, scalable architecture to efficiently handle high-volume file processing with fault tolerance and monitoring in place.
• Built the integration in a configuration-driven manner, enabling dynamic handling of changes such as new files, fields, or schema updates without requiring code changes.
• This approach significantly reduced delivery timelines for enhancements and maintenance, enabling faster releases and improving adaptability to evolving business needs.

Bi-directional Marketo Unsubscribe Sync via Custom Salesforce API:

• Designed, Developed and exposed a Salesforce API with two endpoints to synchronize unsubscribe events between two Marketo instances.
• Each Marketo instance triggers its respective endpoint upon a user unsubscribe, which then updates the other Marketo instance via API calls to maintain consistent unsubscribe status.
• Configured Marketo webhooks to notify Salesforce in real-time.
• Designed a custom Salesforce object to track unsubscribe status and sync state, capturing success and failure details.
• Implemented a batch job and a queuable job in Salesforce for retrying failed syncs and enabling near real-time synchronization.
• Added a manual override checkbox on the Salesforce object to allow triggering of failed syncs on demand.

E.ON (Energy), (UK):

• Integration between SmapOne Cloud to SAP. Periodically fetching data from SmapOne and storing it in on the Azure Storage (Storing in Queue and then dequeuing was involved too).
• Created CI/CD Pipeline for building, running MUnit test case, storing artifacts and deploying the APIs. Used Git, Maven, Jenkins, JFrog, Azure DevOps.

Neural Semantic Search using Sinequa:

• Leveraged Sinequa platform and scheduled jobs to neuralize and index knowledge documents—including Confluence, Salesforce KB, and product documentation—into a vector database.
• Evaluated multiple Hugging Face vectorizer models based on dimensions, parameters, and multilingual capabilities to generate optimized document embeddings.
• Tuned indexing parameters such as chunk size and weighting/boosting to ensure optimal data representation in the vector database.
• Developed APIs to enable semantic search: user queries are converted to vectors, matched using cosine similarity against the vector database, and top results are re-ranked by a Large Language Model (LLM) before being returned with relevance scores.