Repetitive
work sucks.
I automate it.

1. 01 · AskMyGuru

   AI Quality · Automation

   Active

   Can a system read a PRD and a Figma file and produce test cases a human would have written?

   Turning repetitive test design into a generation pipeline with human review.

   Test design per feature, including human review

   1–1.5 days→3–4 hours

   Python / Anthropic Claude SDK / Notion / Figma

   Read the case study →

2. 02 · AskMyGuru

   AI Quality

   Active

   How do you test a system when there isn't always one correct answer?

   Measuring correctness in a system whose output is different every time.

   What the suite could detect

   Functional tests passing→Grounding failures surfaced

   Python / DeepEval / Anthropic Claude SDK

   Read the case study →

3. 03 · Mobile Premier League

   Performance · Reliability

   How do you know a system survives a peak you can't rehearse?

   Validating capacity for traffic peaks that only happen once a year.

   Fantasy traffic before a peak event

   Unverified peak capacity→~1M RPM validated

   Locust / Kubernetes / Prometheus / Grafana / Chaos Mesh

   Read the case study →

Problems I've automated away

1. AskMyGuru

   09/2025 — Present

   Bengaluru, India

   Current

   Senior SDET, Automation and Performance

   Seed-stage AI startup

   • Built a Python and DeepEval evaluation framework for an AI chatbot, using Anthropic Claude as an LLM judge to validate factual accuracy, hallucinations, grounding, relevance and contextual consistency — surfacing grounding and factual-accuracy failures that functional testing did not catch.
   • Built an internal test-case generation framework from PRD/Notion and Figma inputs, later extended with HLD/LLD context, reducing test-case preparation from ~1–1.5 days to ~3–4 hours including human review.
   • Built a human-in-the-loop AI support agent using BM25 retrieval and the Anthropic Claude SDK over 890 historical support replies, grounding draft responses in SOPs before agent review.
   • Built a Python, pytest and Appium mobile automation framework for critical payment, subscription and chat journeys, reducing regression time from ~3 days to ~1 day.
   • Built API automation across backend services for P0/P1 and end-to-end business flows, scaling coverage from 0 to ~1,000 tests within 2 months.
   • Owned performance testing across backend services, using Prometheus and Grafana to monitor latency, throughput, errors and service behaviour under load.

   • Python
   • DeepEval
   • Anthropic Claude SDK
   • pytest
   • Appium
   • Prometheus
   • Grafana

2. Mobile Premier League

   11/2018 — 09/2025

   Bengaluru, India

   Senior Manager, Automation and Performance

   SDET Lead → SDET Manager → Performance SDE Manager → Senior Manager

   • Led a 25+ member QA organisation across Automation, Performance Engineering, Games and Product QA.
   • Owned Fantasy and Payments performance readiness for IPL and other peak events, validating up to ~1M RPM overall Fantasy traffic and up to ~60 payment TPS.
   • Drove performance validation for the AWS-to-GCP migration, covering load testing and database latency investigations.
   • Ran chaos testing on GKE using Chaos Mesh, covering pod kill/failure scenarios and HTTP latency/error injection.
   • Scaled automation from ~200–250 to nearly 2,000 tests, helping shorten the release cycle from ~3 weeks to ~2.5 weeks.
   • Performance-tested the SQS/SNS-to-Confluent Kafka migration, benchmarking broker/node behaviour and cluster capacity under load.
   • Benchmarked G1GC vs ZGC on Java 8/17, comparing GC pauses, throughput and CPU/memory usage under load.
   • Validated EC2-to-EKS/Kubernetes migration performance by benchmarking latency, CPU and memory utilisation under load.
   • Developed backend automated tests using Java and RestAssured for REST APIs and gRPC service-to-service flows not exposed through client-facing APIs.

   • Java
   • Python
   • Locust
   • Kubernetes
   • Chaos Mesh
   • Kafka
   • RestAssured
   • Appium
   • AWS
   • GCP

3. Bookshare · RnR Datalex · BizSense Solutions

   2014 — 2018

   US contract / India

   QA and Automation Engineer

   • Built Selenium and Java regression tests for a US accessibility platform using Page Object Model, TestNG, Maven and Jenkins.
   • Performed functional and automation testing across ERP, e-commerce, healthcare procurement and B2B trading products for US and Dubai clients.

   • Selenium
   • Java
   • TestNG
   • Maven
   • Jenkins

What I work with

AI / LLM Quality

• DeepEval
• LLM-as-judge
• Anthropic Claude SDK
• MCP
• BM25 retrieval
• FastAPI
• Grounding checks
• Hallucination detection

Languages & Test Automation

• Python
• Java
• Bash
• Appium
• Selenium
• Playwright
• pytest
• TestNG
• RestAssured
• REST / gRPC
• Contract testing
• Allure

Performance & Reliability

• Locust
• Load / stress / spike / soak
• JVM & GC benchmarking
• Capacity planning
• WireMock
• Chaos Mesh

Cloud, Containers & CI/CD

• AWS (EKS, EC2, S3, IAM)
• GCP (GKE, Cloud DNS)
• Kubernetes
• Docker
• Jenkins
• GitHub Actions
• Git

Observability

• Prometheus
• Grafana
• Datadog
• New Relic
• OpenTelemetry
• Loki
• Jaeger
• Sentry
• Elasticsearch

Data, Messaging & Caching

• PostgreSQL
• Confluent Kafka
• SQS / SNS
• Zookeeper
• Redis
• ElastiCache
• Memorystore
• Metabase

Currently building

AI-Assisted Test Failure Analysis

Can an AI system understand why an automated test failed, rather than only reporting that it failed?

EXPERIMENT / Active