Trustworthy AI • LLM Testing • Full-Stack Automation

I am an AI Developer and Test Engineer (12+ years) working at the intersection of software engineering and AI assurance. My research-oriented work focuses on converting probabilistic model behavior into repeatable experiments and measurable evidence—so AI-enabled systems remain reliable under change and auditable in production.

I design and build evaluation and automation systems that move beyond “does it work?” to answer: is it safe, stable, and defensible under real-world conditions?

• LLM Testing Hub
  🔗 https://github.com/K11-Software-Solutions/llm-testing-hub
  A research hub for LLM assurance—evaluation harnesses, regression suites, red-teaming scenarios, and reliability scorecards for repeatable, audit-ready testing.

• AI Assurance (Promptfoo)
  🔗 https://github.com/kavitaj11/weoptimize.ai_assurance_promptfoo
  Promptfoo-based AI assurance suite for LLM evaluation, regression testing, red-teaming scenarios, and audit-ready reporting.

• K11 TechLab — Playwright + Python AI-Assisted Automation Framework
  🔗 https://github.com/K11-Software-Solutions/k11techlab-playwright-python-ai-assisted-framework
  Modern Playwright + PyTest automation framework with reusable fixtures, scalable structure, and AI-assisted testing patterns (prompt-driven test generation, resilient execution hooks, and extensibility for LLM/MCP-based workflows).

• K11 Tech Lab — Selenium + Java Full-Stack Framework (AI-assisted)
  🔗 https://github.com/K11-Software-Solutions/k11TechLab-selenium-java-fullstack-framework
  Enterprise automation framework emphasizing modular design, reporting, and repeatable execution—extended with AI-assisted testing and AI integration patterns (prompt-driven test generation, self-healing strategies, and evaluation hooks for AI/LLM-enabled workflows).

• K11 Tech Lab — Cucumber BDD + Java Full-Stack Framework
  🔗 https://github.com/K11-Software-Solutions/k11TechLab-cucumber-bdd-java-fullstack-framework
  BDD framework for scalable automation across UI + API + Mobile workflows, built for CI/CD and long-term maintainability.

• Kavita Jadhav — AI Developer Portfolio (Gemini + Cloud Run)
  🔗 https://github.com/kavitaj11/kavita-jadhav-ai-developer-portfolio
  AI-powered portfolio showcasing full-stack engineering with a Gemini-backed chat experience, secure key handling via backend proxy, Dockerized deployment, and Google Cloud Run production hosting.
  Walkthrough: https://dev.to/kavitaj11/ai-powered-portfolio-full-stack-developer-showcase-with-google-gemini-cloud-run-2aaj

• Full-Stack Developer Capstone
  🔗 https://github.com/kavitaj11/xrwvm-fullstack_developer_capstone
  End-to-end full-stack delivery demonstrating application development, API integration, and deployment-ready structure.

• Apollonia Dental Practice — Employee Management CRUD Web App
  🔗 https://github.com/kavitaj11/apollonia-dental-practice-employee-management-system-CRUD-web-app
  Full-stack CRUD application demonstrating end-to-end delivery.

• Greenspot Grocer — Relational Database Transformation Project
  🔗 https://github.com/kavitaj11/greenspot-grocery-portfolio-database-project
  Data modeling, normalization, and scalable database design from flat datasets.

• Rumi Press — Book Distribution Expense Tracker (Django)
  🔗 https://github.com/kavitaj11/django_based_rumi_press_book_distribution_expense_tracker
  Django-based expense tracking app for book distribution workflows, supporting structured record management, reporting-friendly data organization, and deployment-ready project setup.

• Euro Orbit Travel Agency — 7-Day Weather Forecast App
  🔗 https://github.com/kavitaj11/euro-orbit-travel-agency-7-day-weather-forecast-app
  Weather forecast application providing a 7-day view with API-driven data fetching, user-friendly UI, and deployment-ready structure.

My current interests focus on reliability, safety, and governance-aligned evaluation for LLM-enabled systems:

• LLM evaluation methodology & benchmarking
  • Reproducible harness design, multi-model comparisons, metric selection, regression protocols
• Failure mode characterization
  • Hallucination patterns, prompt sensitivity, bias/fairness gaps, context-window limits, behavioral drift
• Adversarial testing & red teaming
  • Prompt injection, jailbreak resilience, adversarial formatting/obfuscation, privacy/PII leakage, unsafe content pathways
• Robustness & production monitoring
  • Drift detection signals, stability under distribution shift, reliability scorecards, audit-friendly evidence trails
• Evaluation in CI/CD
  • Automated quality gates, traceable results, compliance-aligned test suites, reproducible reporting
• AI integration in legacy test automation frameworks
  • Adding AI-assisted capabilities to existing Selenium/BDD/API frameworks (test generation, self-healing, intelligent assertions, flaky-test reduction) while maintaining determinism, traceability, and maintainable architecture
• AI application development & end-to-end validation
  • Building and testing AI-enabled applications (RAG, agents/tool use, MCP-based integrations), with emphasis on observability, data/privacy controls, and evaluation-driven development

• Evaluation harnesses and test suites for accuracy, faithfulness, safety, and compliance
• Full-stack validation across UI + API + DB integrated into CI/CD pipelines
• Reliability artifacts: scorecards, benchmark runs, failure-mode matrices, audit-friendly reporting
• Frameworks that prioritize testability-by-design and production realism

Banking • Healthcare Insurance (ORMB & EDI workflows) • Financial Trading • SaaS Subscription Commerce (Offer-to-Cash) • iEN Infrastructure Service Assurance

• 🌐 K11 Software Solutions: https://k11softwaresolutions.com
• GitHub: https://github.com/kavitaj11

llm-testing ai-assurance red-teaming prompt-injection jailbreaks evaluation test-automation full-stack ci-cd quality-engineering trustworthy-ai