Amulya Gupta — Work Experience & AI Systems Career

Senior Software Engineer

HCLTech

• Ran 40+ in-app campaigns through Adobe Journey Optimizer during Black Friday and Cyber Monday, reaching 20M+ users with zero critical failures — outcomes that required careful audience filter validation and staged rollouts, not just hitting publish
• Drove 10-15% improvement in CTR and purchase completion through structured A/B testing: each test varied one dimension at a time (subject line length, CTA copy, send timing, segment definition) so the signal was clean and the learnings carried forward to the next campaign
• Built the data pipelines that feed the personalisation engine: raw event ingestion from multiple upstream sources, transformation layer to normalise schemas, and a delivery layer that populates campaign attributes in near real-time — all orchestrated so a failed upstream event doesn't silently corrupt downstream targeting
• Designed the production readiness process for each campaign release: pre-flight checklists covering audience filter accuracy, experience rendering across device types, and content slot population per locale; every release went through a staging smoke test before promotion to production
• Debugged silent failures in personalisation rules — cases where a rule evaluated correctly in isolation but produced wrong results against live audience data due to stale segment membership caches; traced these through pipeline logs and fixed the invalidation logic
• Worked across engineering, QA, and product to formalise the release checklist into a repeatable process, which cut last-minute campaign holds from a regular occurrence to rare exceptions
• Wrote internal documentation on campaign pipeline architecture and failure modes so the process wasn't dependent on any one person — useful when onboarding new team members during peak season prep

Software Engineer I

Classplus

• Worked on decomposing a monolithic backend into independently deployable microservices — the actual work involved mapping inter-service dependencies, deciding ownership boundaries for shared data, and writing migration scripts that kept both the old and new code paths functional during the cutover period
• Chose service boundaries based on team ownership and data coupling, not just technical convenience; services that shared a database table stayed together until we had time to split the schema properly rather than creating distributed monolith anti-patterns
• Built ETL pipelines on MySQL and PostgreSQL processing daily student activity, assignment completion, and engagement data for the ops team's dashboards — pipelines that ran on a schedule, handled partial failures gracefully, and logged row-level errors so bad records could be investigated without rerunning the whole job
• Dashboard query performance improved significantly after adding materialised summary tables updated incrementally by the ETL jobs — queries that previously timed out on large cohorts completed in under 3 seconds on equivalent data volumes
• Wrote backend services in Python and Node.js, picking the language based on the workload: Python for data-heavy analytics endpoints where pandas integration saved time, Node.js for high-concurrency event-driven routes where async I/O mattered more
• Connected REST APIs to frontend reporting views so teachers and school admins could see near real-time student performance data — built the pagination and filtering logic on the API layer so the frontend didn't have to pull full datasets and slice them client-side
• Debugged production data inconsistencies by tracing requests end-to-end through logs, identifying cases where out-of-order event writes caused incorrect aggregation in the reporting layer, and fixing the write ordering guarantees upstream
• Got comfortable with the full cycle of shipping backend features: spec, implement, write integration tests against a test database, deploy to staging, verify with the ops team, promote to production

M.Tech in Artificial Intelligence & Machine Learning (WILP)

Birla Institute of Technology and Science (BITS), Pilani

Coursework: Distributed Systems, Deep Learning, Data Management for ML

M.Sc. General Studies (Data Science Minor)

Birla Institute of Technology and Science (BITS), Pilani