Reliability
Idempotent writes, structured retries, and error handling that survives the worst day.
Enterprise Integration Engineer
Building Reliable Systems. Connecting What Matters.
Production paths · B2B & APIs · Systems that survive Monday morning
Enterprise integrations on MuleSoft and Workato — production APIs, B2B/EDI, and automation for regulated industries and global logistics. I optimize for clarity under load, not slide-deck polish.
I build systems that don’t break in production.
MuleSoft · Salesforce · Workato — vector marks on a dark tile with brand glow (MuleSoft & Salesforce via Simple Icons; Workato from your reference). Scroll tilts the row.
01 About
What I optimize for
I build integration systems that stay calm when traffic spikes, partners change formats, and someone asks “what happened to that message?” long after stand-up ended.
Hyderabad · MuleSoft & Workato · 3+ years shipping production integrations
I care about the boring parts that keep teams out of crisis mode: contracts the next engineer can reason about, queues that don’t swallow failures, and logs you can trace without guesswork. Most of my work sits between “the business needs this data” and “the platform cannot go dark.”
Today I’m with Prowess, on large-scale partner messaging for global shipping. Before that, years at k-logix meant owning client-facing delivery across banking, commerce, and healthcare — migrations, APIs, and automation where a defect has a dollar sign or a human name attached to it.
For dates, titles, and stack-level detail, see Career and Case studies below — this section is the why, not the inventory.
When I take on a system, I take on the responsibility for it — not just the next ticket.
Scalability
APIs and recipes that scale with traffic — not with cost or engineering toil.
Clean API Design
RAML contracts as the source of truth — versioned, documented, consumer-aligned.
B2B / EDI
Partner file and message pipelines: SFTP intake, IBM MQ async paths, EDIFACT-style validation, outbound send, and reconciliation at bulk scale.
02 Expertise
Platforms and depth.
Integration Platforms
- MuleSoft Anypoint
- Mule 4.6.x
- Anypoint Studio
- CloudHub 2.0
- Workato
- Anypoint MQ
- EDI / B2B
APIs & Languages
- REST
- SOAP
- RAML
- DataWeave
- JSON
- XML
- Webhooks
Connectors & Systems
- Salesforce
- Acumatica
- Shopify
- AxisCare
- Snowflake
- Oracle (JDBC)
- IBM MQ
- Database
- SFTP
- HTTP
Testing & DevOps
- MUnit
- SoapUI
- Postman
- Anypoint Monitoring
- Jenkins
- Azure DevOps
- Maven
- Docker
Tools
- GitHub
- Bitbucket
- JIRA
- Confluence
- Slack
Certifications
- MuleSoft MCD Level 1
- Workato Pro 1
- Workato Pro 2
- Workato Pro 3
03 Career
Roles and ownership.
-
Current Integration Developer
June 2025 — Present
- EDI/B2B on Mule 4.6.x for PIL (Pacific International Lines) — multi-partner pipelines, queue-backed processing, map/send flows, and bulk-scenario reconciliation alongside performance validation.
- Owned DataWeave and flow logic for partner message transformation, validation, control-number handling, and error paths that match shipping operations.
- Integrated SFTP, IBM MQ, HTTP/REST system APIs (including JDBC to Oracle DB), wired for async scale and traceability across the pipeline.
- Supported SIT and non-prod performance cycles — structured payloads, cold-start vs steady-state runs, worker CPU/Java heap, queue depth, and Anypoint Monitoring signals.
- CI-backed delivery with peer review, version control on GitHub, and daily Agile alignment with the team.
-
Integration Developer
Dec 2022 — May 2025
- Owned end-to-end delivery of MuleSoft and Workato integrations across banking, commerce, and healthcare clients — from RAML design to CloudHub deployment and post-prod support.
- Architected and executed CloudHub 1.0 → 2.0 migration for production banking APIs with zero customer-facing incidents.
- Cut Workato task consumption through deliberate recipe redesigns — removing redundant triggers and replacing polling with webhooks.
- Hardened error handling and observability on critical flows — structured retries, dead-letter routing, and audit trails that turn 2 AM pages into 5-minute fixes.
- Established testing standards (MUnit + SoapUI) that caught regressions before SIT — and made releases boring, in the best way.
- Partnered directly with U.S.-based clients in daily Agile ceremonies — owning requirements, design, and delivery without a layer in between.
-
B.Tech, Computer Science & Engineering
2018 — 2022 · Dean's List · GPA 6.53/10
- Distributed systems, web technologies, and software engineering fundamentals — the base I still lean on when designing integrations and failure modes.
04 Case studies
Evidence, not adjectives.
Each write-up is short on fluff: problem, approach, outcome. Expand any card for detail.
★ Flagship
MuleSoft
June 2025 — Present
PIL (Pacific International Lines) — B2B EDI Integration
Global shipping partner messaging at scale: inbound files and industry B2B messages from SFTP through IBM MQ–backed async processing, DataWeave map and validation, to outbound send and downstream handoff — API-led, production-disciplined.
Problem
Multiple trading partners, bulk file volumes, and strict B2B semantics — without losing messages, without blind spots in reconciliation, and without performance surprises when queues and workers are under load.
Solution
Contributed to API-led Mule 4.6.x services across inbound ingestion, transformation/map, and outbound send: SFTP sources, IBM MQ for scalable async processing, DataWeave for partner-specific formats and validation (including control-number / code-list handling), HTTP/REST system APIs (including JDBC to Oracle), structured error paths, and SIT validation with scenario-based payloads.
Impact
- End-to-end traceability: inbound files → consumed messages → transformed records → successful sends vs failures — reconciled for bulk test runs
- Non-production performance baselines: cold-start vs steady-state, throughput, queue depth, worker CPU/Java heap — using Anypoint Monitoring
- Defects caught and closed in SIT before they could surface as production incidents
- Peer-reviewed, CI-gated merges — predictable releases instead of hero fixes
- Mule 4.6.x
- DataWeave
- SFTP
- IBM MQ
- HTTP
- Oracle JDBC
- Anypoint Monitoring
- MUnit
- Maven
- GitHub
MuleSoft
Dec 2022 — May 2025
BayFirst National Bank — Enterprise Integration Estate
A U.S. community bank running mission-critical APIs on aging CloudHub 1.0. I owned the migration, the contracts, and the production handoff — end to end.
Problem
The bank's MuleSoft estate was on CloudHub 1.0 with inconsistent error handling, tight coupling, and limited observability. Deprecation was on the roadmap; downtime was not an option.
Solution
Owned the CloudHub 2.0 migration end-to-end — re-architected flows around Anypoint MQ, redesigned RAML contracts with explicit versioning, built MUnit coverage, and rolled out via phased deployment with rollback gates.
Impact
- Zero customer-facing incidents across the entire migration window
- Right-sized vCore + worker config — lower runtime cost on the same throughput
- Versioned RAML contracts removed an entire class of consumer-side regressions
- Production issues now triaged from logs in minutes, not hours
- Owned the run-book — no escalations to other teams during cutover
- Mule 4.x
- RAML
- DataWeave
- Anypoint MQ
- SFTP
- Salesforce
- MUnit
- SoapUI
Workato
Aug — Oct 2024
Puffco — Commerce ↔ ERP ↔ CRM Sync
Manual reconciliation between Acumatica, Shopify, and Salesforce was slowing fulfillment and creating data drift. I designed the automation that fixed it.
Problem
Order, customer, and inventory data lived in three systems that didn't talk. Ops spent hours daily reconciling Shopify orders against Acumatica fulfillment and Salesforce account state — and still missed things.
Solution
Designed event-driven Workato recipes with webhook ingestion from Shopify, structured payload mapping to Acumatica, and Salesforce sync flows with audit logging and clean error handling.
Impact
- Replaced overnight batch reconciliation with real-time sync
- Order, customer, and inventory state now consistent across all three systems
- Eliminated the daily reconciliation workload from ops
- Failed payloads land in a dead-letter queue with full context — not lost
- Workato
- Salesforce
- Acumatica
- Shopify
- Webhooks
- JSON
Workato
Oct — Dec 2024
Arosa Care — Multi-System Client Data Automation
A home-care provider with client and sales data fragmented across Salesforce, AxisCare, and Snowflake. I built the automation that made all three trustworthy.
Problem
Client records were duplicated across systems, sales opportunities went stale, and downstream reporting in Snowflake couldn't be trusted — making it hard to act on pipeline data.
Solution
Built a Workato automation layer with query-list driven deduplication, real-time webhook updates between Salesforce and AxisCare, and an audit trail that traced every record change back to source.
Impact
- Duplicate client records eliminated at the source — not patched downstream
- Workato task consumption cut through deliberate trigger and polling redesign
- Snowflake reporting trustworthy again — sales now act on pipeline data
- Every record change traceable to source via the audit layer
- Workato
- Salesforce
- AxisCare
- Snowflake
- Webhooks
- HTTP
05 Playground
Route Rush
Payloads fall in three lanes. When a dot crosses the purple glow band, tap Left / Center / Right for that lane — or keys 1 2 3. Route before it hits the bottom. Three misses and the backlog wins.
Score 0
Best 0
Lives 3
06 Contact
Let's build something
worth shipping.
Have an integration that needs to scale — or one that’s breaking in production? Let’s talk.