Skip to content
View MukundaKatta's full-sized avatar
🎯
We taught silicon to speak. Now it’s asking who it is. So am I.
🎯
We taught silicon to speak. Now it’s asking who it is. So am I.
9 followers · 3 following

Achievements

Achievement: Pair Extraordinairex4Achievement: Galaxy Brainx2Achievement: Pull Sharkx3Achievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pair Extraordinairex4Achievement: Galaxy Brainx2Achievement: Pull Sharkx3Achievement: YOLOAchievement: Quickdraw

Block or report MukundaKatta

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
MukundaKatta/README.md
Mukunda Rao Katta

Portfolio LinkedIn X Email


I ship production AI / ML systems at Fortune-100 scale and contribute upstream to the SDKs that power them.
Currently at Southwest Airlines · formerly Software Engineer at AWS.

Now: building agent guardrails, MCP linters, and eval harnesses; contributing to AI SDKs upstream.


Impact

Production AI / ML
COST
−78%
SageMaker → Bedrock
$1,740 → $371 / mo 		LATENCY
600×
retrieval improvement
aircraft fault prediction 		RAG SCALE
30K+
knowledge entries
9-stage agentic pipeline 		QUALITY
370+
tests & evaluations
across production systems

Open Source Footprint
ORIGINALS
164
public repos
I authored 		PACKAGES
50
npm + PyPI + MCP
across 3 registries 		UPSTREAM
98
merged PRs in
external public repos 		REPOS REACHED
409
external projects
I've contributed to

Selected Work

📦 streamparse

npm · CLI · Homebrew · MCP Registry · TypeScript · 64 tests · 0 deps

Streaming JSON parser that yields partial valid trees as tokens arrive. Render LLM tool calls mid-stream, recover dropped responses, parse messy ```json blocks. Single-pass state machine with synthetic closure so every snapshot is valid JSON.

npm MCP Registry

🛡️ mcpcheck

npm · GitHub Action · SARIF · TypeScript · CLI · CI-ready

Lint MCP config files for Claude Desktop, Cursor, Cline, Windsurf, and Zed. Catches malformed transports, missing auth, duplicate servers, and placeholder values. Drops into any CI as a 3-line composite Action with SARIF output for code-scanning.

npm downloads

🧪 ai-eval-forge

npm + PyPI · zero deps · CLI + programmatic API

Eval harness for comparing model, prompt, and agent behavior. Exact, regex, token-F1, JSON, and citation-coverage checks. CLI for one-shot runs, programmatic API for CI integration. Sister Python release on PyPI.

npm PyPI

📸 agentsnap

npm · zero deps · 33 tests · TypeScript types

Snapshot tests for AI agents — record an agent run's tool-call trace, diff against a baseline, fail CI on regressions. Drops into any test runner. First in a four-package agent reliability stack (snap · guard · vet · cast · fit).

npm
+ 46 more packages across npm, PyPI, the MCP Registry, GitHub Marketplace Actions, and Homebrew

Distribution pattern. Each flagship ships across the surfaces it makes sense to live on:

library  →  Python port  →  CLI binary  →  GitHub Action  →  Homebrew formula  →  MCP server
   npm           PyPI                          Marketplace        brew tap            npm

So the same problem is solvable from a TypeScript app, a Python script, a CI workflow, a terminal, or directly inside Claude / Cursor / Cline / Windsurf / Zed.

Categories (npm + PyPI):

Area Examples
Structured outputs & parsing streamparse, streamparse-mcp, llm-response-schema-lite
Agent reliability agentsnap, agentguard, agentvet, agentcast, agentfit
Agent infrastructure agent-loop-breaker, agent-regression-lens, tool-call-contracts, tool-permission-gate
RAG & retrieval rag-quality-kit, retrieval-acl-filter, vector-poison-score, embedding-dedupe
Prompt & output safety pii-sentry, prompt-injection-shield, llm-output-sanitizer, system-prompt-leak-scan
Cost, routing, caching llm-cost-guard, model-fallback-planner, model-router-policy, semantic-cache-key
MCP & Claude Code linting mcpcheck, mcp-config-check, claude-skill-check, claude-hooks-check, claude-commands-check
Evals & tracing ai-eval-forge, eval-dataset-smith, eval-flake-detector, llm-trace-sampler, agent-run-diff

Browse the full set at npmjs.com/~mukundakatta and pypi.org/user/mukundakatta.

GitHub Marketplace Actions: mcpcheck, claude-skill-check, mcp-config-check, claude-hooks-check, claude-commands-check.

Homebrew tap: brew tap mukundakatta/tools then any of the linters above.


Recent Open Source

Selected upstream contributions, recent first:

Where I tend to land: AI SDKs, MCP tooling, eval frameworks, agent infrastructure, structured outputs. Public log of selected work in oss-contributions.


Recently Shipped

Last refreshed 2026-04-26 from npm, PyPI, and the GitHub API.

Latest releases

Recently merged PRs


Featured Projects

Karna — AI Agent Platform

Self-hosted AI assistant across 7 messaging channels (Telegram, Slack, Discord, WhatsApp, SMS, iMessage, Web). Extensible plugin SDK, semantic memory, voice. TypeScript monorepo with Next.js dashboard and React Native mobile app.

TypeScript · Next.js · Supabase · WebSocket · pgvector

Chetana — AI Consciousness Research

Research platform for studying machine consciousness via 14 indicators grounded in 6 scientific theories. Turns abstract questions into structured experiments, scoring, and analysis.

Python · Evaluation · Experimentation

AgentRAG — Modular RAG

Provider-agnostic RAG framework with pluggable vector stores, chunking strategies, and retrieval methods. Designed for agentic workflows with clean API boundaries.

TypeScript · Vector Search · Embeddings

Astra Agent — Runtime

Standalone AI agent runtime with tool execution, context management, and multi-model routing. Foundation for autonomous AI assistants with structured tool use.

TypeScript · LLM Orchestration · Tool Use
+ more projects (Sadhak, Prithvi, AgentMem, Patchly, Evalharness, TokenWise, …)
Project What it does
Sadhak AI-powered job search command center — automated evaluation, resume tailoring, application tracking
Prithvi Container security scanner — vulnerability detection, compliance checks, Docker audits
AgentMem Pluggable memory management for AI agents
Patchly AI code review bot — flags bugs, suggests fixes, explains why
Evalharness Prompt, agent, and RAG test harness — red teaming, regression testing
TokenWise Token usage optimization across providers
LLM Bench CLI Benchmark local LLMs — speed, throughput, quality
Amogha Cafe Full-stack Firebase restaurant platform · live

Experience

Role Company Era
AI/ML Engineer Southwest Airlines Aug 2025 — Present
AI/ML Engineer GPS IT Solutions Jun 2024 — Aug 2025
Software Development Engineer Amazon Web Services Aug 2022 — May 2024
Data Engineer GPS IT Solutions Jan 2022 — Aug 2022
Software Engineer American Express Feb 2017 — Dec 2020
Highlights from each role

Southwest Airlines — AI/ML Engineer

  • Architected ML fault-prediction system for aircraft maintenance: 5 prediction types, 10K+ records, sub-second retrieval.
  • Led SageMaker → Bedrock migration: 78% cost reduction ($1,740 → $371/mo), 600× latency improvement.
  • Designed 9-stage agentic RAG pipeline (LangGraph, Bedrock Nova Pro/Micro, FAISS + BM25) over 30K+ KB entries.

GPS IT Solutions — AI/ML Engineer

  • Built GPT-4 + RAG content generation platform with compliance validation; reduced production time by 40%.
  • Designed AI model-risk governance framework with 23 automated evaluation tests (regulatory compliance).
  • Architected FastAPI microservices with FAISS / Pinecone vector search on Kubernetes.

Amazon Web Services — SDE

  • Shipped features for AWS Application Manager (Systems Manager) serving enterprise customers globally.
  • Owned full-stack delivery: React/TypeScript frontend + Java/Python backend APIs.
  • Designed CI/CD and IaC patterns enabling zero-downtime deployments at enterprise scale.

GPS IT Solutions — Data Engineer

  • Led end-to-end migration of data pipelines from on-prem to AWS (Glue, PySpark).

American Express — Software Engineer

  • Built Python backend services and REST APIs for enterprise platforms handling high-volume transactions.

Stack

What I reach for, in order of frequency:

TypeScript · Python · React/Next.js · Node.js · FastAPI · AWS (Bedrock, SageMaker, ECS, OpenSearch) · LangGraph · Postgres + pgvector · Docker / Kubernetes · Terraform


GitHub Activity

STARS
1395 		COMMITS
7400 		PRS
993 		CONTRIBUTIONS
8703 		REPOS REACHED
409

stars cumulative · commits & contributions are last 12 months · PRs all-time · repos reached = external repos contributed to





Education & Credentials

M.S., Big Data Analytics & Information Technology · University of Central Missouri (2022) B.Tech, Mechanical Engineering · SRM University (2016)

Selected certifications

Anthropic — MCP Advanced Topics · Claude with Bedrock · Claude with Vertex AI · Building with Claude API · Intro to Agent Skills · Intro to Subagents

AWS — Generative AI Applications · Services for AI Solutions · AI Fundamentals & the Cloud · Amazon Q for Software Dev

Cloud / GCP — Advanced Terraform with GCP · Build & Deploy Agent with Reasoning Engine

Stanford — Machine Learning · Introduction to Statistics

Wharton (Coursera) — Business Analytics · Customer Analytics · People Analytics

Microsoft — Generative AI for Software Devs · GitHub Copilot Productivity


Open to opportunities

Senior AI/ML Engineer · GenAI Platform Engineer · Software Engineer

Las Vegas, NV · mukunda-ai.vercel.app


Profile Views GitHub Followers

Pinned Loading

  1. karna karna Public

    Karna — Your Loyal AI Agent Platform. Self-hosted personal AI assistant with multi-channel messaging, extensible skills, and semantic memory.

    TypeScript 2

  2. rnht rnht Public

    Rudra Narayana Hindu Temple — community platform for events, donations, and priest scheduling

    TypeScript 1

  3. amogha-cafe amogha-cafe Public

    Amogha Cafe & Restaurant — Premium full-stack Firebase web app with ordering, admin dashboard, kitchen display, QR dine-in, live tracking & more. Live at https://amogha-cafe.web.app

    JavaScript 1

  4. astra-agent astra-agent Public

    Standalone AI agent runtime — tool execution, context management, and multi-model routing for autonomous assistants

    TypeScript 1

  5. chetana chetana Public

    Chetana — AI Consciousness Research Platform. Test AI models against 14 consciousness indicators from 6 scientific theories.

    TypeScript 1