Pavan Aditya Chillara

Led techno-functional enhancements on Infor XA finance streams; mapped business rules into modular RPG components and parameter-driven SQL procedures for rapid adaptability. Designed microservice-style boundaries on IBM i by isolating domain services (posting, inquiry, allocations) behind IWS endpoints with clear versioning and backward compatibility. Implemented event-driven integration using data queues and change-data capture (CDC) queries; published lightweight events consumed by external services for near-real-time updates. Optimized quarter-close pipeline with parallelized batch chains (dependency graph in control tables), savepoints, and idempotent step restarts to withstand partial failures. Built a JSON validation gateway (YAJL + custom rules) to quarantine malformed requests; added replay tooling with diff view to compare original vs. corrected payloads. Enforced data governance through row-level masking in views for PII, authority collection reviews, and standardized GRTOBJAUT scripts in deploy packs. Extended stored procedures to support window functions and analytical aggregates for allocations and aging; replaced spool-heavy reports with SQL views consumed by BI. Introduced a Git-backed source workflow on IBM i through VS Code (Code for IBM i); implemented pre-commit RPG syntax checks and SQL formatting to enhance code quality. Created observability hooks—timing probes in RPG and SQL Service metrics (QSYS2 services) feeding a lightweight dashboard for SLA and capacity monitoring. Drove zero-downtime deployments for APIs by coordinating IWS instance pools, rolling promotes via MDCMS, and using config toggles to shift traffic post health-check validation. Stabilized production through targeted index maintenance windows, proactive RUNSTATS equivalents, and performance watch queries to detect regressions ahead of cutoffs.

Engineered high-throughput posting services using ILE RPG with service programs and bound APIs; standardized prototypes in copy members to enforce interface contracts. Modernized integrations by wrapping core RPG logic behind IWS REST services emitting/consuming JSON; utilized YAJL for serialization and schema validation. Tuned long-running finance jobs through SQL query rewrites, multi-column indexes, and optimizer hints (where justified), achieving >40% runtime reduction via selective materialized results. Implemented auditable posting with full journal capture, commit scopes, and idempotent replays; introduced sequence tokens to prevent duplicate postings during retries. Authored SQL stored procedures and functions for FX revaluation, intercompany eliminations, and accrual reversals; leveraged common table expressions (CTEs) for clarity and performance. Built robust subfile workflows for dispute/adjustment handling (row-level actions, soft deletes, reason codes) with strict authority checks via adopted profiles. Established a unified error taxonomy (business vs. system) and centralized exception logging with correlation IDs; integrated alerts to ServiceNow with enriched context payloads. Migrated DDS files to DDL, eliminating redundant LFs in favor of targeted SQL indexes; documented access-path strategies and rebuild policies. Packaged promotions via MDCMS with pre-/post-deploy SQL and authority scripts; implemented automated rollback plans and four-eyes approval enforcement. Containerized API contract tests (mock callers) outside IBM i; executed in Azure DevOps pipelines post-promote to validate request/response fidelity. Built performance dashboards (SQL views + queries) for finance close—tracking job durations, row counts, and blockers per period to support capacity planning.

Refactored legacy monolithic RPG into an ILE modular design using service programs and binding directories; decomposed business rules into procedures to enable unit-level testing and reuse. Implemented embedded SQL with cursors, row-level triggers, and MERGE statements for upsert logic; standardized SQLSTATE handling and diagnostics logging for consistency. Designed DDL-based tables, views, and indexes for finance subledgers; added constraints and foreign key cascades to prevent orphan balances and ensure referential integrity. Built robust Subfile UIs with expand/collapse, position-to, and filtered search capabilities for reconciliation screens; optimized record formats to reduce display I/O overhead. Externalized configuration (company, plant, currency, GL mapping) into control tables with effective-dated logic to support back-valued postings. Created stored procedures for period-close checks (open items, suspense GL, FX revaluation) and wrapped them in RPG orchestrators to drive end-to-end batch flows. Introduced YAJL (JSON) parsing for inbound payloads from upstream applications; validated schemas and converted them into internal structures for posting engines. Developed IWS (Integrated Web Services) endpoints to expose ledger inquiry functions to downstream systems; secured endpoints with basic authentication, TLS, and request throttling. Hardened error handling with MONITOR/ON-ERROR blocks and dead-letter tables; built replay utilities to reprocess failed transactions safely. Migrated change control to MDCMS across dev/test/prod, enabling deploy sets, approvals, automated object promotions, and consistent naming/versioning policies. Built SQL-based health checks (aging, data drift, orphan checks) scheduled nightly; surfaced aggregated results via a dashboard query for support team monitoring.

Built batch programs in free-form RPGLE and CLLE to automate daily postings, end-of-day reconciliations, and file rollovers; parameterized CL commands to support multiple company codes without duplication. Modeled core data structures in DB2 for i using DDS (PF/LF) and migrated high-volume reads to SQL DDL tables with proper unique keys, reducing logical rebuild overhead. Developed interactive screens with Display Files and Subfiles (paged and load-all variants); implemented function-key handling and message subfiles for user feedback and error display. Replaced OPNQRYF and ad-hoc QRY objects with embedded SQL; optimized filters to be SARGable and added selective indexes to eliminate full table scans. Introduced commitment control (COMMIT/ROLLBACK) for financial postings; configured journal receivers and ensured transaction safety across multi-file updates. Wrote stored procedures (SQL/PSM) encapsulating reusable posting and validation logic; exposed them to RPG programs via prototypes for type-safe invocation. Created Printer Files for statutory and audit reports; implemented selective reprint and spool archival following standardized OUTQ conventions. Automated housekeeping tasks (cleanup, archiving, report bundling) via CLLE jobs scheduled in JOBSCDE, with outcomes logged into control tables. Used STRDBG (with update capability) to step through RPG procedures; analyzed DSPJOBLOG for CPF/SQL diagnostics and resolved data integrity issues through controlled fix scripts. Authored detailed data mapping and program flow documentation, including copybook/prototype catalogs and dependency diagrams for batch processing chains.