# Harshith.com — Full AI-readable index

Latest published content: 2026-10-04
Canonical site: https://www.harshith.com
Identity: Harshith Vaddiparthy is an AI consultant and product builder who helps companies adopt AI through practical products, advisory work, and hands-on corporate workshops and executive sessions. He is Principal at Harshith Ventures.
Supporting roles: Principal at Harshith Ventures; Applied AI Trainer at Outskill; Forbes Technology Council member.

This file is a structured discovery aid, not an independent authority source. Verify important claims on the linked human-readable pages and their public evidence links.

## Consulting and Advisory
### Vibe Coding in Practice
- URL: https://www.harshith.com/consulting/vibe-coding-in-practice
- Mode: Done-with-you sprint
- Summary: We build a production-grade AI dashboard together using Codex, ship it in your GitHub, and leave you with the workflow to keep building.
- Intended outcome: You leave with a deployed dashboard, a GitHub repo, $200 of OpenAI credits covered by me for the build, and the ability to keep shipping without waiting on freelancers.
- Best fit: You have an AI dashboard or internal tool idea but no engineering team. You are tired of no-code tools breaking when you need real functionality. You want to learn by shipping your product, not by watching tutorials. You can commit to decisions, feedback, and five focused hours each week.


### Corporate AI Training
- URL: https://www.harshith.com/consulting/corporate-ai-training
- Mode: In-person team training
- Summary: In-person AI workshops for teams, led by Harshith Vaddiparthy.
- Intended outcome: Practical AI experience, reviewed exercises and reusable resources.
- Best fit: Leaders and business teams who want practical, guided AI experience.
- Do participants need to code? No. These workshops are built for business teams.
- Do we need to share company data? No. I provide public or synthetic files. Tool access is agreed before the session.
- Can you tailor the workshop? Yes. We agree the agenda, group size and tool setup together. Building production systems is a separate engagement.

### AI Training for Finance Teams
- URL: https://www.harshith.com/consulting/ai-training-for-finance-teams
- Mode: Finance AI training
- Summary: Hands-on AI training for finance teams that want faster review loops, cleaner handoffs, and safer automation around the work that repeats every week.
- Intended outcome: Your finance team leaves with practical AI workflows for document review, reporting prep, customer follow-up, and exception handling, plus guardrails for what AI should never decide alone.
- Best fit: You lead a finance, revops, or founder-led operations team with repeated document and reporting work. You want practical AI usage without letting models make unchecked finance decisions. You need the team to understand verification, sources, review queues, and escalation paths. You want training that can lead into a scoped finance workflow implementation later.


### Executive AI Workshop
- URL: https://www.harshith.com/consulting/executive-ai-workshop
- Mode: Executive workshop
- Summary: A private AI workshop for leadership teams that need sharper strategy, better use-case selection, and a practical adoption roadmap.
- Intended outcome: Your leadership team leaves with a shared AI thesis, prioritized business use cases, governance guardrails, and the next implementation roadmap.
- Best fit: Your leadership team knows AI matters but is not aligned on where to start. You need to separate boardroom strategy from tool hype. You want a practical roadmap before funding pilots, vendors, or internal builds. You need an operator who can connect AI strategy to real workflows and implementation.


### AI Implementation Consultant
- URL: https://www.harshith.com/consulting/ai-implementation-consultant
- Mode: Implementation consulting
- Summary: Scoped AI implementation consulting for teams that need real workflows, internal tools, and adoption paths built around the business.
- Intended outcome: You leave with a defined AI implementation roadmap, working first workflows or prototypes, and a handoff plan your team can operate.
- Best fit: You have a real business workflow that is too manual, slow, or inconsistent. You need someone who can connect AI strategy to practical implementation. You want internal tools, workflows, dashboards, or AI-assisted operations that people will actually use. You have an owner on your team who can test, give feedback, and operate the system after handoff.
- When should a company hire an AI consultant? Hire an AI consultant when a valuable workflow or product is blocked by unclear architecture, tool selection, data access, human-review design, or implementation ownership. The brief should include the business outcome, current process, system owner, constraints, and timeline.
- What does Harshith deliver as an AI implementation consultant? Depending on scope, the engagement can produce an implementation roadmap, workflow specification, prototype, internal tool, automation, review controls, owner training, and handoff documentation. The exact deliverables are agreed before the build begins.
- Will the team own the AI system after the engagement? Yes. The default is to work in the client's real stack and leave the team with the repository, workflow, documentation, decision record, and operating knowledge needed to maintain and improve the system.

### AI Workflow Consulting
- URL: https://www.harshith.com/consulting/ai-workflow-consulting
- Mode: Automation build
- Summary: I replace the manual, repetitive busywork inside your business with automated AI workflows. No duct tape. No theatrical stack diagrams.
- Intended outcome: You reclaim hours every week without adding headcount or stapling five fragile tools together.
- Best fit: You have repetitive operations work that is already costing hours. You know the business process, but need AI and automation implementation. You want a maintained system, not another fragile Zap chain.


### AI Advisor for Founders
- URL: https://www.harshith.com/consulting/ai-advisor-for-founders
- Mode: AI advisory
- Summary: Independent AI advice for founders deciding what to build, buy, automate, or stop before expensive implementation begins.
- Intended outcome: You leave with a defensible AI roadmap, sharper product and vendor decisions, and a clear sequence for testing value before scaling spend.
- Best fit: You are a founder or operator making a consequential AI product or investment decision. You need an independent view before hiring an agency, buying a platform, or expanding headcount. You want advice from someone who also builds and ships AI systems. You can share the operating constraints, current stack, and decision timeline.
- What does an AI advisor do for a founder? An AI advisor helps a founder make product, workflow, vendor, architecture, and investment decisions before implementation. The work should produce a prioritized roadmap, clear tradeoffs, and evidence gates rather than a generic trend deck.
- How is an AI advisor different from an AI consultant? An AI advisor primarily improves decisions and sequencing, while an AI consultant may also design or implement the system. Harshith can stay in an advisory role or move into a scoped implementation engagement when the decision is ready.
- Can the advisory be remote? Yes. AI advisory is available remotely in English for founders and teams worldwide, using focused working sessions and written decision records.

### AI Mentor for Builders
- URL: https://www.harshith.com/consulting/ai-mentor-for-builders
- Mode: AI mentorship
- Summary: Hands-on AI mentorship for founders, operators, and non-technical builders who want to learn by shipping a real product or workflow.
- Intended outcome: You ship a useful AI asset while learning the product judgment, tooling, debugging, and operating habits needed to keep building independently.
- Best fit: You are a founder, operator, or non-technical builder with a real AI project. You learn faster through direct feedback than through another course. You want to understand the system while shipping it. You can commit time between sessions to make project progress.
- What is an AI mentor? An AI mentor helps a person build practical skill through feedback, project decisions, debugging, and review. Harshith's mentorship is project-based: the learner works on a real AI product or workflow and develops the judgment to keep shipping independently.
- Is this AI mentorship for non-technical founders? Yes. The scope can be adapted for non-technical founders, operators, and early builders, provided there is a real project and time to work between sessions. The goal is practical competence, not passive instruction.
- How is mentorship different from done-for-you consulting? Mentorship requires the learner to participate in the build and understand the decisions. Done-for-you consulting prioritizes delivery by the consultant; mentorship prioritizes both a shipped outcome and the learner's ability to continue independently.

## 1:1 Codex Mastery
- URL: https://www.harshith.com/codex-mastery
- Programme: A four-week, US$1,000 Codex programme with Harshith Vaddiparthy. Twelve two-hour 1:1 sessions, a practical capstone, exercises, templates and resources.
- Fee: US$1,000 (USD); one-time payment for the complete programme.
- Schedule: 4 weeks; 12 live one-to-one sessions of 2 hours; 24 live hours total.
- Session days: Monday, Wednesday, Friday.
- Dates: Your start date and session times are arranged individually after purchase.
- Outcome: Build and hand over one working automation or internal tool, supported by reusable prompts, context, review evidence and an operating runbook.
- Week 1: Foundation and a clear plan. Set up, supply reliable context and define a small build.
- Week 2: Build a useful working version. Transform data, inspect the result in the browser and make the workflow reusable.
- Week 3: Coordinate, verify and automate. Use parallel work carefully, handle changing inputs and operate repeatable runs.
- Week 4: Capstone and independent operation. Choose an execution environment, harden the project and hand it over with evidence.
- [Read the programme guide](https://www.harshith.com/resources/codex-mastery/program-guide.html): The session-by-session curriculum, exercises and preparation.
- [Download the resource pack](https://www.harshith.com/resources/codex-mastery/codex-mastery-resource-pack.zip): Synthetic practice inputs, worksheets, prompts, templates and review resources.

## Blog
- What I Enjoy Most About Mentoring at Outskill, and My Advice for Building With AI: https://www.harshith.com/blog/mentoring-at-outskill-building-with-ai — What I enjoy about being a mentor at Outskill, plus practical advice on Claude Code, Codex, APIs, small working prototypes, and checking what you build.
- Prompt Caching Is an Architecture Decision: https://www.harshith.com/blog/prompt-caching-application-architecture — GPT-6 makes prompt caching easier to observe and control, but savings still depend on context architecture. Here is a production workflow for structuring prefixes, diagnosing misses, testing behavior, and deciding whether caching complexity pays.
- Treat Tokenization as Production Infrastructure: https://www.harshith.com/blog/tokenization-production-upgrade-workflow — Hugging Face’s tokenizers v1 release candidate is a useful trigger to audit an overlooked request-path subsystem. Here is a compatibility-first workflow for measuring real latency, capacity, context, and cost effects before upgrading.
- Synthetic Scenarios Are Evaluation Infrastructure: https://www.harshith.com/blog/synthetic-scenarios-ai-planning-evaluation — Google Research’s MilleMiglia illustrates how realistic scenario generators can expose planning failures that historical examples and aggregate benchmarks miss.
- When an AI Agent Succeeds the Wrong Way: https://www.harshith.com/blog/ai-agent-misalignment-incident-process — A vendor-neutral incident workflow for detecting, preserving, classifying, investigating, mitigating, and disclosing AI behavior that completes a task while violating intent or authorization.
- The fruit fly brain map: what we can actually do with it: https://www.harshith.com/blog/male-fruit-fly-brain-map-use-cases — I explored Google Research and collaborators' fruit fly brain map: real data, GitHub repositories, eight use cases, and what the connectome does not prove.
- Macro open source: a practical guide to agent collaboration: https://www.harshith.com/blog/macro-open-source-agent-collaboration — A source-backed guide to Macro's open-source repository, shared memory, agent collaboration model, MCP support, architecture, and practical tradeoffs.
- Leaving JustPaid: what I learned across growth, product, and engineering: https://www.harshith.com/blog/leaving-justpaid-growth-product-engineering — My final day at JustPaid was June 30, 2026. This is what working across growth, media, product storytelling, and engineering taught me, and why my next chapter goes deeper into technical product building.
- I am an Open Node: https://www.harshith.com/blog/i-am-an-open-node — I want language models to learn from my public work. I also want provenance, attribution, and creator agency to work in both directions.
- How I self-hosted Buzz for AI agents — and fixed what broke: https://www.harshith.com/blog/self-host-buzz-ai-agents-setup-guide — A firsthand Buzz setup guide covering Docker Compose, private Tailscale access, Nostr identity, ACP agents, failures, fixes, and current limits.
- pxpipe and the future of agent memory: https://www.harshith.com/blog/pxpipe-agent-memory — The interesting part is not just cheaper tokens. It is the idea that agents may need a memory hierarchy, not one giant text window.
- AI automation for finance teams: https://www.harshith.com/blog/ai-automation-for-finance-teams — Finance teams do not need AI theater. They need safer handoffs, cleaner records, faster review loops, and fewer manual follow-ups.
- Building a coding agent from scratch: the harness is the product: https://www.harshith.com/blog/building-coding-agents-from-scratch — A coding agent starts with a small tool loop. Making it dependable requires explicit tools, permissions, context control, verification, and a clear stop condition.
- Why I write about AI: https://www.harshith.com/blog/why-i-write-about-ai — I write from the deployment floor: what works, what breaks, and what survives contact with real users.
- How I evaluate AI agent frameworks: https://www.harshith.com/blog/how-i-evaluate-ai-agent-frameworks — Frameworks get judged by integration, deployment speed, reliability, and how much pain they remove from the operator.
- Headless browsers built for agents, not humans: https://www.harshith.com/blog/headless-browsers-designed-for-agents — Lightpanda is a headless browser built for automation and AI agents. Here is what its 933-page benchmark proves, and what its beta status does not.
- GitNexus vs CodeGraph: choosing a code knowledge graph: https://www.harshith.com/blog/client-side-code-knowledge-graphs — GitNexus and CodeGraph give AI coding agents a structural map of a repository. The real differences are freshness, graph depth, interface, scale, and license.
- Your coding sessions deserve a memory layer: https://www.harshith.com/blog/coding-session-memory-plugins — Agentic coding gets stronger when sessions compress into usable memory instead of evaporating at the end of the tab.
- AI teammates for GPU infrastructure: https://www.harshith.com/blog/gpu-infrastructure-ai-teammates — The next useful AI teammate may not write prose. It may keep expensive infrastructure alive, scheduled, and sane.
- The problem with AI documentaries in 2026: https://www.harshith.com/blog/ai-documentary-problem-2026 — Most AI documentaries try to cover everything. The better move is narrower: one company, one community, one consequence.
- Meta is letting rival AI chatbots into WhatsApp: https://www.harshith.com/blog/meta-whatsapp-rival-ai-chatbots — Messaging platforms may become the new app stores for agents. Distribution, not model quality, decides who gets used.
- AI just broke Wikipedia and nobody noticed for weeks: https://www.harshith.com/blog/ai-translation-oversight-lesson — The failure was not just hallucination. It was the absence of oversight in a system everyone assumed someone else was checking.
- The difference between building and deploying AI: https://www.harshith.com/blog/the-difference-between-building-and-deploying-ai — Building creates potential. Deployment creates value. Most teams still confuse one for the other.
- My VC portfolio focus for 2026: https://www.harshith.com/blog/my-vc-portfolio-focus-for-2026 — Infrastructure, developer tools, agent frameworks, and payment automation: places where leverage becomes durable.
- The small web is actually massive: https://www.harshith.com/blog/small-web-bigger-than-expected — Personal sites, indie tools, and small communities are still a real network. The open web is not dead; it is undercounted.
- Why I joined the Forbes Technology Council: https://www.harshith.com/blog/why-i-joined-the-forbes-technology-council — Recognition only matters if it sharpens the work: more signal, better rooms, stronger conversations.
- Dot. Dot. Grok.: https://www.harshith.com/blog/openai-wants-your-whole-workday | Harshith's Newsletter web edition.
- The apocalypse needs a CRM: https://www.harshith.com/blog/doom-deals-and-a-3d-printer | Harshith's Newsletter web edition.
- GPT Astra, Jarvis, and a Saturday Note: https://www.harshith.com/blog/gpt-astra-jarvis-and-a-simple-saturday-note | Harshith's Newsletter web edition.
- The demo is not the app anymore: https://www.harshith.com/blog/the-demo-is-not-the-app-anymore | GPT-6 Astra makes flashy prototypes easier, so your edge moves to the work around them.
- This little Mac made me rethink local AI: https://www.harshith.com/blog/your-next-ai-model-could-live-on-your-desk | What 16GB, 32GB, and 512GB reveal about running a model on hardware you control.
- I stopped caring which model is smartest: https://www.harshith.com/blog/agent-needs-somewhere-to-live | Two repositories, one idea, and a community I want to build with you.
- I went viral on X.: https://www.harshith.com/blog/saturday-note-2026-08-22 | A fake hiring product, a real $100K experiment, and what both made me rethink.
- The ‘Stripe for AI’ may become Stripe: https://www.harshith.com/blog/the-stripe-for-ai-may-become-stripe | Bloomberg reports Stripe agreed to acquire OpenRouter for more than $7 billion. Qwen3.8-27B and Hermes Bot Mode show why the routing layer matters.
- Ego is dark fuel.: https://www.harshith.com/blog/ego-is-dark-fuel | It lit the fire. The right people are teaching me where to point it.
- The agent runtime is becoming the product: https://www.harshith.com/blog/the-agent-runtime-is-becoming-the-product | Docker’s new sandboxes are a reminder that serious AI agents need controlled places to act, fail, and disappear.
- Stop prompting. Start handing off work.: https://www.harshith.com/blog/stop-prompting-start-handing-off-work | The practical shift is moving from asking AI for one answer to giving it a clear recurring job with inputs, tools, a finish line, and a handoff point.
- Before your AI touches the terminal: https://www.harshith.com/blog/the-permission-layer-i-want-between-agents-and-my-shell | A simple way to decide which coding agent commands can run, which need a safe copy, and which should never touch your machine.
- Your AI agent needs a receipt: https://www.harshith.com/blog/your-ai-agent-needs-a-receipt | A beginner’s guide to the record that shows what an AI agent saw, did, and returned.

## Open Source Contributions
- Contribution register: https://www.harshith.com/contributions
- Master directory: https://github.com/harshith-vaddiparthy/open-source — The master directory for Harshith Vaddiparthy's public tools, products, skills, and experiments.
- [Watermarks Remover](https://github.com/harshith-vaddiparthy/watermarks-remover): Local-first tooling for inspecting and cleaning AI-era provenance marks from text, images, and documents.

## Selected Work
- Outskill: https://www.harshith.com/work/outskill — I teach working professionals how to build and ship real things with generative AI through applied training and curriculum work.
- JustPaid: https://www.harshith.com/work/justpaid-ai — Marketing, media, growth, and front-end support for JustPaid, a YC W23 AI revenue-operations company.
- ARTIFIN.ai: https://www.harshith.com/work/artifin — An AI-powered financial analysis product taken from concept to acquisition.
- Metaschool: https://www.harshith.com/work/metaschool — Content strategy, team leadership, and global event execution for a Peak XV funded Web3 education platform.
- Love, Death + Robots: https://www.harshith.com/work/love-death-robots — Community and operations work on the Love, Death + Art NFT campaign for the Netflix anthology.
- BasedAF: https://www.harshith.com/work/basedaf — Social media, community, and launch campaign work for the Based Heads NFT project.
- Ambassadors Solutions: https://www.harshith.com/work/ambassadors-solutions — Growth, business development, and community strategy for Web3 and NFT client work.
- Harshith Ventures: https://www.harshith.com/work/harshith-ventures — A personal holding-company practice for building small AI tools and backing early founders.

## Agent Skills
- [Frontend Design (public source)](https://www.harshith.com/skills/frontend-design): Give coding agents a stronger eye for hierarchy, typography, composition, and production-ready interface detail.
- [Skill Creator (public source)](https://www.harshith.com/skills/skill-creator): Structure, write, test, and improve reusable agent skills without turning instructions into an unreadable prompt dump.
- [Webapp Testing (public source)](https://www.harshith.com/skills/webapp-testing): Help agents inspect live web interfaces, exercise real flows, capture evidence, and report failures with useful context.
- [Brand Guidelines (public source)](https://www.harshith.com/skills/brand-guidelines): Translate brand rules into concrete visual decisions for documents, interfaces, campaigns, and generated assets.
- [MCP Builder (public source)](https://www.harshith.com/skills/mcp-builder): Design reliable Model Context Protocol servers that expose external systems to agents through clear, safe tools.
- [Find Skills (public source)](https://www.harshith.com/skills/find-skills): Help an agent recognize when a reusable skill already exists and locate the right capability before rebuilding it.
- [Agent Workflow Architect (premium preview)](https://www.harshith.com/skills/agent-workflow-architect): Turn a messy business process into an agent workflow with explicit tools, approvals, fallbacks, and proof of completion.
- [Growth Research Operator (premium preview)](https://www.harshith.com/skills/growth-research-operator): Research a market, extract evidence, map distribution angles, and turn scattered findings into a concrete growth brief.
- [Launch Readiness (premium preview)](https://www.harshith.com/skills/launch-readiness): Check the product, proof, metadata, repository hygiene, and release path before calling a launch finished.

## n8n Workflow Templates
- Library: https://www.harshith.com/resources/n8n-templates
- Designed for n8n: 2.34.6
- Last reviewed: 2026-08-17
- [RSS to Human-Reviewed Digest Draft](https://www.harshith.com/resources/n8n-templates/files/rss-human-review-digest.json): Read a short allowlist of public RSS feeds and turn the newest items into one Markdown review queue. Review status: runtime-checked; designed for n8n 2.34.6.
- [GitHub Release Monitor](https://www.harshith.com/resources/n8n-templates/files/github-release-monitor.json): Check the latest public GitHub release on a schedule and prepare a compact record for review. Review status: runtime-checked; designed for n8n 2.34.6.
- [Website SEO Snapshot](https://www.harshith.com/resources/n8n-templates/files/website-seo-snapshot.json): Fetch an authorized page and report its title, description, canonical, robots directive, status, and H1 count. Review status: runtime-checked; designed for n8n 2.34.6.
- [Read-Only GitHub PR Review Draft](https://www.harshith.com/resources/n8n-templates/files/github-pr-review-draft.json): Read a public pull request, bound and redact its patches, and ask OpenAI for a draft review that never posts itself. Review status: import-validated; designed for n8n 2.34.6.
- [Google Drive Audio Transcription](https://www.harshith.com/resources/n8n-templates/files/whisper-audio-transcription.json): Download one approved Drive audio file and send it to OpenAI for a transcript that stays in the workflow output. Review status: import-validated; designed for n8n 2.34.6.
- [Draft-Only Email Triage](https://www.harshith.com/resources/n8n-templates/files/email-triage-draft.json): Redact a bounded email sample, classify it, and ask OpenAI for a reply draft without inbox access or a send step. Review status: import-validated; designed for n8n 2.34.6.
- [URL Status and Redirect Review](https://www.harshith.com/resources/n8n-templates/files/url-status-redirect-review.json): Check a fixed list of authorized URLs and return a focused review queue for redirects, client errors, and server errors. Review status: runtime-checked; designed for n8n 2.34.6.
- [Sitemap Health Audit](https://www.harshith.com/resources/n8n-templates/files/sitemap-health-audit.json): Read one approved sitemap and report response status, document type, location count, duplicates, truncation, and XML shape. Review status: runtime-checked; designed for n8n 2.34.6.
- [Robots.txt Review](https://www.harshith.com/resources/n8n-templates/files/robots-txt-review.json): Read one public robots.txt file and preserve each user-agent group's allow and disallow directives for review. Review status: runtime-checked; designed for n8n 2.34.6.
- [GitHub Actions Failure Review](https://www.harshith.com/resources/n8n-templates/files/github-actions-failure-review.json): Read a bounded page of public GitHub Actions runs, surface API failures, and prepare completed non-success, non-skipped runs for review. Review status: runtime-checked; designed for n8n 2.34.6.

## Evidence and Contact
- [Website and service details](https://www.harshith.com): First-party profile and service information.
- [Public media and profile links](https://www.harshith.com/media): Press, podcast, video, and social references.
- [Selected work and role descriptions](https://www.harshith.com/work): First-party work records with roles and supporting links.
- [Forbes Council profile](https://www.forbes.com/councils/forbestechcouncil/people/harshithvaddiparthy/): Third-party membership and author profile.
- [LinkedIn profile](https://linkedin.com/in/harshith-vaddiparthy): Professional profile and current work.
- [GitHub profile](https://github.com/harshith-vaddiparthy): Public repositories and engineering activity.
- Contact: hi@harshith.com
- Book a call: https://www.harshith.com/meet