GEOPROBE / WEBSITE DIAGNOSTICS

A clearer view of amankushwaha.dev.

Your findings, the evidence behind them, and a practical next step.

Findings & evidence

Sorted by severity, then weight

Mentioned on third-party sites AI trusts

15 pts

No mentions of “Aman Kushwaha” on 5 sites AI cites most (Wikipedia, Product Hunt, GitHub, Hacker News…).

Recommended next step

AI answers are built from a small set of trusted sites. In order of impact for you:
- Wikipedia — up to half of ChatGPT's top-10 citations
- Product Hunt — launch page AI uses to confirm a product exists
- GitHub — top-30 overall; decisive for developer tools
- Hacker News — developer and startup discussion
- Stack Overflow — developer Q&A

View evidence

brand: Aman Kushwaha
found: none
missing: Wikipedia, Product Hunt, GitHub, Hacker News, Stack Overflow
coverage: 0 of 9 weighted points
notChecked: Reddit, YouTube, LinkedIn, G2, Capterra, TrustRadius, Medium, Quora, Trustpilot (needs BRAVE_SEARCH_API_KEY REDDIT_CLIENT_ID/SECRET)

Needs attention

Structured data (JSON-LD)

15 pts~10 min fix

Found Person, WebSite. Missing a page-level type (SoftwareApplication, Product, Article…).

Recommended next step

Add a page-level type describing what this page is about.

View evidence

types: ["Person","WebSite"]

Warning

AI crawlers can access your site

20 pts

All major AI crawlers (OpenAI, Anthropic, Perplexity, Google) are allowed.

View evidence

allowed: ["GPTBot","OAI-SearchBot","ChatGPT-User","ClaudeBot","Claude-User","PerplexityBot","Google-Extended"]
blocked: []
robotsTxt: User-Agent: * Allow: / Disallow: /dashboard Disallow: /studio Disallow: /studio/ Disallow: /api/ Disallow: /api/analytics/ User-Agent: Twitterbot Allow: / User-Agent: facebookexternalhit Allow: / User-Agent: GPTBot Allow: / User-Agent: OAI-SearchBot Allow: / User-Agent: ChatGPT-User Allow: / User-Agent: ClaudeBot Allow: / User-Agent: Claude-User Allow: / User-Agent: anthropic-ai Allow: / User-Agent: PerplexityBot Allow: / User-Agent: Google-Extended Allow: / User-Agent: Applebot-Extended Allow: / User-Agent: CCBot Allow: / User-Agent: Bingbot Allow: / User-Agent: llms.txt Allow: / Host: https://amankushwaha.dev Sitemap: https://amankushwaha.dev/sitemap.xml

Pass

Content is readable without JavaScript

20 pts

4,127 characters of readable text are served in the HTML.

View evidence

visibleChars: 4127
noscriptWarnsJs: false

Pass

llms.txt is present

10 pts

/llms.txt is present (4,692 bytes).

View evidence

bytes: 4692
hasH1: true
hasFull: false
hasLinks: true

Pass

Headline and intro answer "what is this?"

10 pts

H1 and a 21-word intro paragraph are present.

View evidence

h1: I build AI/ML systemsthat reach production.
firstParagraph: I'm Aman Kushwaha, an AI/ML Engineer building retrieval and fine-tuning pipelines — and the fast, accessible web products that ship them.

Pass

Title, description and canonical tags

5 pts

Title, description, canonical and Open Graph tags are all set.

View evidence

title: Aman Kushwaha — AI/ML Engineer & Full Stack Developer
canonical: https://amankushwaha.dev
description: AI/ML engineer and full stack developer building retrieval systems, fine-tuning and evaluation pipelines — and the fast, accessible Next.js products that ship them.

Pass

Content freshness signals

5 pts

Most recent update 35 days ago.

View evidence

datedEntries: 16
newestDaysAgo: 35
sitemapEntries: 23

Pass

Your generated llms.txt

Built from your sitemap and navigation. Review it, then upload to https://amankushwaha.dev/llms.txt

llms.txt37 lines

# Aman Kushwaha

> AI/ML engineer and full stack developer building retrieval systems, fine-tuning and evaluation pipelines — and the fast, accessible Next.js products that ship them.

Website: https://amankushwaha.dev

## Pages

- [Projects](https://amankushwaha.dev/projects): Selected projects by Aman Kushwaha — AI/ML systems, retrieval pipelines, full stack web apps and developer tooling built with Python, Next.js, TypeScript and R…
- [Skills](https://amankushwaha.dev/skills): The technical toolkit of Aman Kushwaha — AI/ML (PyTorch, RAG, fine-tuning, evaluation), frontend (React, Next.js, TypeScript), backend (Node.js, Python) and cl…
- [Resume](https://amankushwaha.dev/resume): Résumé of Aman Kushwaha, AI/ML Engineer — experience, education and skills.
- [VoxRAG](https://amankushwaha.dev/projects/voxrag-voice-rag): VoxRAG is a voice- and text-enabled Retrieval-Augmented Generation system built on MSMARCO-XI. Speak or type in English, हिन्दी, मराठी, বাংলা, മലയാളം, ગુજરાતી,…
- [AAIPL Tournament: Dual-Agent Fine-Tuning Solution](https://amankushwaha.dev/projects/aaipl-tournament-dual-agent-fine-tuning-solution): Won the AAIPL Hackathon (1st place out of 120 teams) by building a dual-agent LLM system using Qwen2.5-14B-Instruct. Developed end-to-end fine-tuning pipelines…
- [The Diary](https://amankushwaha.dev/projects/the-diary): The Diary is a minimal, distraction-free social platform built for busy professionals to stay connected with family and close friends through daily journal ent…
- [AkashSetu](https://amankushwaha.dev/projects/Exo-Planet-AkashSetu): AI-enabled detection of exoplanets from noisy TESS light curves, using a Box-Least-Squares transit search feeding a calibrated XGBoost + Random-Forest ensemble…
- [Genome AI: Cattle Breed Classification from SNP Data](https://amankushwaha.dev/projects/genome-ai-cattle-breed-classification-from-snp-data): A reproducible machine learning pipeline for cattle breed classification from SNP genotype data, featuring CNN and Transformer baselines. Includes data validat…
- [Conveyor Belt Detection](https://amankushwaha.dev/projects/conveyor-belt-detection): Labeled custom conveyor belt data and trained a YOLOv8 model from scratch. Increased detection accuracy from 72% to 80% with model tuning and augmentation.
- [RubricAI](https://amankushwaha.dev/projects/rubricai-smart-academic-evaluation): RubricAI is a powerful Next.js application designed for educators and students to evaluate assignments against custom rubrics using Genkit AI. It features batc…
- [Attendance System Using OpenCV](https://amankushwaha.dev/projects/attendance-system-using-opencv): A Python-based attendance system that utilizes facial recognition to accurately identify students and mark their attendance on specific dates.

## Blog

- [Blog](https://amankushwaha.dev/blog): Writing on web development, system design, AI/ML and software engineering by Aman Kushwaha — tutorials, build notes and lessons learned.
- [Building Deploy Detective, Approval-Gated Incident Response Agent on TrueForge](https://amankushwaha.dev/blog/building-deploy-detective-trueforge-agent): Learn how we built Deploy Detective, an incident response agent that investigates alerts, finds root cause in a sandbox, and pauses for human approval before r…
- [Adaptive Parallel Reasoning: Scaling LLM Inference with Dynamic Parallelism](https://amankushwaha.dev/blog/adaptive-parallel-reasoning-llm-inference): Explore how adaptive parallel reasoning lets LLMs dynamically allocate compute, reducing latency and improving accuracy during inference.
- [Beyond PEFT: Why LoRA May Not Be the Best Fine‑Tuning Choice](https://amankushwaha.dev/blog/beyond-peft-why-lora-may-not-be-the-best-fine-tuning-choice): Explore why PEFT techniques beyond LoRA can offer better trade‑offs in accuracy, memory usage, and runtime for fine‑tuning models.
- [Gradient-Based Planning for Long‑Horizon World Models with GRASP](https://amankushwaha.dev/blog/gradient-based-planning-world-models): Explore how gradient‑based planning (GRASP) makes long‑horizon world‑model control robust and fast.
- [GLM-5.2 Powers Long-Horizon Tasks with 1M Token Context](https://amankushwaha.dev/blog/glm-5-2-powers-long-horizon-tasks-1m-token-context): Discover how GLM-5.2 advances long-horizon tasks with a solid 1M-token context, IndexShare efficiency, and agentic RL improvements.
- [NVIDIA Blackwell Sets New Records in MLPerf Training 6.0 with Unmatched Scale and Performance](https://amankushwaha.dev/blog/nvidia-blackwell-mlperf-training-6-0-records): NVIDIA Blackwell dominates MLPerf Training 6.0, delivering the fastest training time at scale and top per‑accelerator performance.
- [World-Action Models: From Pretrained Imagination to Real‑World Action](https://amankushwaha.dev/blog/world-action-models-pretrained-imagination): Explore how World-Action Models build on vision‑language pretraining to enable robots that can imagine outcomes and act safely.
- [Boosting MoE Training Throughput with Advanced Fusion Kernels](https://amankushwaha.dev/blog/moe-training-throughput-fusion-kernels): Learn how advanced fusion kernels dramatically boost MoE training throughput on GPU clusters.

## Company

- [About](https://amankushwaha.dev/about): About Aman Kushwaha — AI/ML engineer and full stack developer in India, open to remote work. Builds retrieval and fine-tuning pipelines and the web systems tha…
- [Contact](https://amankushwaha.dev/contact): Get in touch with Aman Kushwaha — Open to remote AI/ML and full stack roles. Also available for freelance projects and collaborations.

What to do next

Review the checks that need attention first. Each includes a suggested change and an effort estimate.
Upload the generated llms.txt to your site root.
Come back and hit Rescan to confirm the score moved.