Build
Software, operational tools, data pipelines and automation designed for real workloads.
SOFTWARE · PLATFORM · RELIABILITY
Engineering systems
that survive
production.
Rishabh Gupta — Software, Platform & Reliability Engineering. I work across software, cloud, data and production operations — connecting architecture decisions with the realities of running systems every day.
View engineering work Resume PDF pending
Current focus Cloud reliability · data platforms · AI infrastructure
BUILD → SCALE → OPERATE → EVOLVEEXPLORE THE WORK ↓
01 / THE ENGINEER
Built for the real world.
Architecture matters. So does the 2 a.m. incident, the backfill that needs replaying, and the person inheriting the runbook.
I approach engineering as a lifecycle: design the system, deliver the capability, operate it in production and use what it teaches us to improve the next version.
My work and interests span software engineering, SRE, DevOps, cloud infrastructure, data platforms and AI infrastructure. The through-line is making complex systems more dependable and easier to operate.
That means clear service boundaries, practical automation, useful telemetry, recoverable data flows and decisions that people can explain. I value technical leadership that makes trade-offs visible and leaves teams with better operating context.
Scale
Cloud, distributed processing and platform decisions shaped by throughput and failure modes.
Operate
Observability, incident response and recovery as first-class parts of the design.
Evolve
Modernization, data quality and AI infrastructure without losing operational clarity.
02 / ENGINEERING CAPABILITIES
Depth across
the stack.
Grouped by problems to solve, not logos to display. From runtime infrastructure to the human workflows that keep it working.
Reliability engineering
Move from reactive support toward observable, diagnosable services and repeatable response.
- SRE & production support
- Incident response
- Observability & alerting
- Operational automation
Cloud & platform
Build the deployment and runtime foundations that make delivery safer and operations simpler.
- AWS
- Kubernetes & Docker
- Linux
- CI/CD & Jenkins
Data platforms
Treat correctness, throughput and recoverability as connected design constraints.
- Databricks & Spark
- Kafka & SQL
- Data quality
- Database integration
AI infrastructure
Explore knowledge systems as engineered services with ingestion, retrieval and operational boundaries.
- Local LLMs & Ollama
- Embeddings & Qdrant
- RAG & MCP
- Security & evaluation
03 / SELECTED ENGINEERING WORK
Experience shaped
by operations.
Grouped by engineering context because verified employment dates, titles and individual project ownership are not available for publication.
M3 / adidas context
Production operations
Work spanning operational tooling, health monitoring, incident workflows and data operations. The portfolio focuses on system design and operational approach rather than confidential implementation details.
Selected engineering work
Data & integration
Experience areas include Lambda pipelines, Kafka integration, data quality, SQL databases and production support across M3, MHDR, Claims Highway and SAM contexts. Exact timelines and individual ownership are not asserted here.
Selected engineering work
Service modernization
Modernization and support themes include Angular/Node, SSIS/SSRS, certificate management, L2/L3 investigation and incident coordination with ServiceNow and Opsgenie.
04 / ARCHITECTURE & OPERATIONS
Think in systems.
Operate in loops.
A reference model for connecting application behavior, data movement and the feedback needed to keep both reliable. Illustrative, not a confidential deployment diagram.
REFERENCE PATTERN / PRODUCTION SYSTEM
USERS & OPERATORS
APPLICATION / API controlled actions · service boundaries
DATA PIPELINE ingest · validate · replay
STORAGE & ANALYTICS
OBSERVABILITY metrics · logs · traces
INCIDENT RESPONSE
FEEDBACK → REVISIT DESIGN & OPERATING ASSUMPTIONS
Reliability is a design input.
Define how a service is observed before it fails. Make failure paths debuggable, decide what can be retried safely and keep humans in control of risky actions.
Data flow is an operational contract.
Partitioning, backpressure, idempotency and schema changes affect the downstream system. Track freshness, reconcile results and provide a way to recover incomplete runs.
AI needs an operating model.
For retrieval systems, provenance and permissions matter as much as model quality. Local inference and vector search are explorations here, not claims of production AI deployment.
THE OPERATIONS LOOP
Ship. Observe. Learn. Repeat.
SLIs, SLOs and error budgets are useful when they drive decisions; monitoring, incident response, postmortems and capacity planning complete the loop.
- 01Design
- 02Build
- 03Deploy
- 04Observe
- 05Detect
- 06Respond
- 07Recover
- 08Learn
- 09Improve
05 / PROJECTS / CASE STUDIES
The work behind
the systems.
Technical walkthroughs of problem framing, constraints, architecture, reliability, trade-offs and next steps. Scope is clearly marked on each study.
01 / SRE / OPERATIONSProfessional engineering context · architecture-level overview
M3 operations platform
Operational tooling that connects production health, data workflows and incident investigation.
02 / DATA / PLATFORMEngineering approach · no private pipeline specifications disclosed
Reliable data movement
Designing batch and streaming integration around correctness, throughput and recovery.
03 / AI / EXPERIMENTALPersonal exploration · not presented as a production deployment
Knowledge & AI infrastructure
Exploring local models and retrieval as an auditable knowledge-engineering workflow.
06 / TOOLS & PRINCIPLES
Tools are the means.
Judgment is the skill.
Technologies and practices across professional experience and hands-on exploration. The AI stack is an area of active exploration, not a production credential.
Cloud & runtime
AWS · Lambda · Kubernetes · Docker · Linux · CI/CD · Jenkins
Data systems
Python · SQL · Spark · Databricks · Kafka · MySQL · PostgreSQL · Aurora · MongoDB / DocumentDB
Observability & response
Instana · Grafana · CloudWatch · OpenSearch / Elasticsearch · ServiceNow · Opsgenie
Emerging infrastructure
Bedrock · Ollama · Qdrant · MCP · knowledge ingestion · RAG
ENGINEERING PRINCIPLES
- Design for failure, not just the happy path.
- Observe before optimizing.
- Automate repeatable operational work.
- Make data quality a production concern.
- Keep architecture proportional to the problem.
07 / CONTACT / WHAT'S NEXT
Let's build what
lasts.
I'm focused on reliable cloud and data platforms, distributed systems, AI infrastructure and the operational practices that make them useful. Verified contact and profile links will be published here once supplied.
Contact channels pending verification · No placeholder destinations