Machine Learning Software Engineer

Atul Kumar Tiwary

Boston, MA +1-857-***-****

Email: ******.*@************.*** LinkedIn: linkedin.com/in/tiwaryatul GitHub: github.com/atulty EDUCATION

Northeastern University, Boston, MA Jan 2025 – Apr 2026 M.S. in Computer Science, Khoury College GPA: 3.83/4.0 Relevant coursework: Programming Design Paradigm; Database Design; Algorithms; Web Development; Machine Learning Visvesvaraya Technological University, Belagavi, India Jul 2018 – May 2022 B.E. in Information Science GPA: 8.35/10.0

TECHNICAL KNOWLEDGE

Languages: Java 8/17, Python, C, C++, R

Web UI: JavaScript, TypeScript, Angular, ReactJS, HTML5, CSS3 Frameworks & Tools: Spring Boot, REST APIs & Microservices, Git, Bitbucket, JIRA, Jenkins, Docker, Kubernetes, Liquibase, Autosys, JMeter

Databases & Messaging: Oracle SQL, Cassandra, GemFire, Kafka, IBM MQ Testing & Monitoring: JUnit 5, Cucumber (BDD), Splunk, Grafana, Geneos CERTIFICATIONS

• AWS Certified Cloud Practitioner (Expected Sep 2025) WORK EXPERIENCE

JPMorgan Chase, Bengaluru, India Sep 2023 – Dec 2024 Software Engineer I, Posting & Balance Platform, Cross Border Payments Team (5-engineer team supporting $10B+/day flows)

• Eliminated legacy Lion feed via Spring Boot & Feign REST : cut reconciliation from 1h30m to 40m at 400 TPS.

• Built scalable loan-sweeps ingestion (Oracle & GemFire) : reduced persist latency by 50% at peak 500 TPS.

• Fixed Kafka overflow in production: processed 5k QPS, compressed JSON 10%, meet 1MB limit for 500k+ daily records.

• Persisted 1B+ records/day-GemFire & Oracle via concurrent streaming-reduced ingestion from 50min to 5min-VLS.

• Led Kafka & Cassandra pipelines (Global & India), ingesting 200k msg/s & persisting 10B+ msgs/day in both clusters.

• Resolved DST bug in production : removed UTC GMT conversion, corrected 100K+ timestamps within 2h SLA.

• Developed a Kafka-Cassandra resiliency library for on-demand replay of failed messages, improving recovery time by 80%.

• Implemented HPA on GAIA/EKS scaled balance API pods 3 30 under 200k msg/s, maintaining < 50 ms tail latencies.

• Built Prometheus & Grafana dashboards for balance-service endpoints, cutting MTTD from 15 min to under 5 min.

• Automated Liquibase diff checks between on-prem Oracle & EKS PostgreSQL, catching 100% schema drift pre-production.

• Optimized composite indexes on roll-up tables, reducing average balance-query latency from 200 ms to 20 ms at 5k QPS. Software Developer Consultant, Cross Border Payments Team, JP Morgan - mthree Sep 2022 – Sep 2023

• Mitigated Saudi branch incident : purged 1M+ failed accounts via API and re-triggered scheduler within SLA.

• Migrated Java 8 17 & Spring Boot 2.3 MonetaBoot 3.2, eliminated 20+ bean conflicts, improved startup by 25%.

• Authored Liquibase change-sets for Kafka ingestion DB on GAIA Terraform & AWS EKS, standardizing 200+ table schemas.

• Implemented REST endpoint for multi-currency audit-trail (CoreUI Oracle), handling 500 TPS to support product launch.

• Championed BDD (Cucumber) & JUnit suites covering daily/intraday flows, raising test coverage from 70% to 95%.

• Monitored Splunk, Grafana & Geneos alerts : maintained 99.9% uptime on balance APIs and reduced MTTD by 30%. Northeastern University, Boston, MA

Finance Systems Support Analyst – Internal Consulting Division Jul 2025 – Present

• Managed 1k+ ServiceNow tickets/month with Jagger, Workday & IntelliBuy ERP : achieved 95% SLA compliance.

• Developed runbooks and SQL/Linux log–analysis procedures, reducing ticket resolution time by 30%. PROJECTS

Calendar Application (Java CLI & Swing GUI app for CS5010) GitHub Mar 2025 – Apr 2025

• Implemented interval-tree conflict detection & resolution, preventing overlaps with O(log n) scheduling, 20 ms lookup.

• Enabled CSV import/export & Google Calendar sync for 10k+ daily/weekly/monthly recurring events via runtime flags.

• Designed Swing GUI with event grid for 10k+ entries, achieving sub-100 ms rendering and boosting user efficiency by 30 %. ArrayMiner 4 (Data-driven web app (team of 2) for analytics workflows) Spring 2026

• Designed end-to-end data pipeline handling 100K+ records/day; React dashboards for metrics & MongoDB for database.

• Built REST APIs with Express Mongoose; documented specs user guides for rapid onboarding. LEADERSHIP INTERESTS

• CS3200 TA JPMC CFG Hackathon SME: Mentored 40+ students, co-designed labs/exams, UNIX SME at CFG.

• Interests: Web development, travel, cricket (JP Morgan GLCM Premier League 2024 champion).

Contact this candidate