---
title: "Harshith Nayaka L - AI Engineer in Bengaluru | Tal"
description: "Meet Harshith Nayaka L, AI Engineer, Full-Stack — LLM agents, RAG pipelines, multi-modal systems, production-grade output reliability | AI Workflow Engineer @ DemandNXT | Bengaluru based in Bengaluru. AI stack includes aistudio, anthropic apis, and claude code. Explore public work links including g..."
image: "https://tal.club/meet/harshith-nayaka-l/share-image"
---

# Harshith Nayaka L on Tal

## Headline

AI Engineer, Full\-Stack — LLM agents, RAG pipelines, multi\-modal systems, production\-grade output reliability \| AI Workflow Engineer @ DemandNXT \| Bengaluru

## About

I build AI applications and systems — the engineering that turns a language model from an impressive demo into something a product can actually depend on\. Based in Bengaluru \(Bangalore\), India\. I've worked with LLMs since 2022\. Most of my work lives in the gap between "the model gave an answer" and "the system did the right thing reliably\." That gap is where the real work is: orchestration, structured outputs, validation, and handling the ways models fail\. What I build: → AI agents & orchestration — multi\-step, tool\-using systems that route by cost and capability, and recover from failure → RAG & retrieval — pipelines grounded in real data, validated before anything reaches a user → Multi\-modal pipelines — voice, vision, and text in one coherent flow → Prompt engineering for reliability — reasoning pathways, not hardcoded answers, with model\-based QA before output is used → Full\-stack AI apps — React/Node with AI wired in as a dependable component → Workflow automation — n8n and integration pipelines for shipping AI into real operations Recent work, live and open\-source: → Maestro — multi\-model orchestration engine: a conductor routes one task across thinker/worker/verifier roles and returns a verified answer with a full decision\-log → Cannon — multi\-agent assistant where each domain is an independent specialist, isolation enforced at the query rather than by convention → ReplyDesk — WhatsApp lead agent replying in \~8 seconds with a live ops dashboard Semi\-finalist at the Google Cloud Gen AI Exchange Hackathon 2025 \(270,000\+ developers\)\. Published in IJRTMR on local ML inference — DOI 10\.59256/ijrtmr\.20250506023\. Building something with LLMs that needs to work the hundredth time, not just the first? Send me what you're trying to do and I'll tell you how I'd structure it\. Portfolio: harshith\-nayaka\-l\-portfolio\.vercel\.app

## Location

Bengaluru

## AI stack

aistudio

anthropic apis

claude code

claude

codex

openai

## Work experience

AI Engineer at DemandNXT, Sep 2026 \- Present · 1 mo, Bengaluru, I build Al applications and systems end\-to\-end \- from understanding the business problem and designing the architecture to building the frontend, backend, Al layer, integrations, evaluation, and production workflows\. Based in Bengaluru, India, I currently work as an Al Engineer, handling Al initiatives across 6 B2B brands\. My work spans full\-stack Al products, internal tools, agentic systems, dashboards, Al\-powered marketing operations, SEO/AEO/GEO, and workflow infrastructure\. I've worked with LLMs since 2022\. Most of my work lives in the gap between "the model produced an answer" and "the system reliably produced the right outcome\." What I build: → Full\-stack Al apps \- frontend, backend, model/API integrations, dashboards, deployment, and production workflows Al agents & orchestration \- multi\-step, tool\-using systems with routing, verification, and failure recovery → AI Search / AEO / GEO \- machine readability, Al discoverability, retrieval, structured evidence, and brand visibility in Al answers → RAG & retrieval \- grounding models in real data with validation before information reaches users → Multimodal Al \- text, vision, documents, and generated media inside larger product workflows → Evaluation & reliability \- structured outputs, deterministic checks, model\-based QA, and failure handling → Al\-powered operations \- systems for content, social media, email campaigns, marketing, and business processes Recent projects: → SPECTRA \- Al\-search visibility platform measuring Al readiness and real brand visibility using custom\-intent queries and search\-grounded evaluation → Maestro \- multi\-agent orchestration with Conductor, Thinker, Worker, Verifier, and Synthesizer roles → Personal OS MCP \- an MCP capability layer connecting Al clients with tools and resources on my machines → BrandForge \- end\-to\-end Al creative pipeline with brand intelligence, generation, multimodal QA, and selective repair Semi\-finalist at the Google Cloud Gen Al Exchange Hackathon 2025 \(270,000\+ developers\)\. Published research on local ML inference in IJRTMR\. I'm interested in where Al stops being a demo and becomes products and systems people can actually depend on\. Portfolio: https://harshith\-nayaka\-l\-portfolio\.vercel\.app GitHub: https://github\.com/HarshithNayakaL

AI Workflow Engineer at DemandNXT, Mar 2026 \- Sep 2026 · 6 mo, Bengaluru, Built and deployed AI\-powered workflows, internal tools, and automation systems supporting multiple B2B brands at DemandNXT\. • Designed end\-to\-end AI workflows integrating LLMs, APIs, business data, and internal systems using n8n, Python, Node\.js, and custom integrations\. • Built AI\-powered content and social media systems covering content generation, humanization, brand\-specific copy, visual generation, validation, and operational workflows\. • Developed custom frontend interfaces and backend services around AI workflows, turning internal automations into usable tools rather than standalone workflows\. • Built automation and tooling for email campaign operations, helping campaign teams with data processing, content generation, validation, and repetitive operational tasks\. • Worked on SEO and website automation across multiple B2B brands, including technical SEO, AI\-oriented search optimization, website performance, and WordPress operations\. • Built internal dashboards and command centers integrating multiple APIs and operational data sources\. • Containerized and deployed workflow infrastructure using Docker and built supporting services for local and production\-style execution\. This role evolved from primarily building AI workflows into broader ownership of full\-stack AI products and systems, leading to my transition into an AI Engineer role\.

Data Science Intern at KodNest, Feb 2026 \- Jul 2026 · 6 mos, Bengaluru, Curriculum internship completed as part of my BE final\-semester \(VTU\) requirement\. Structured program building full\-stack and AI applications end to end\. \- Built and deployed Nova AI — an LLM chat application \(Meta Llama 3\.1 8B via Hugging Face Inference API\) with a serverless proxy that keeps the API token server\-side instead of exposing it in the frontend \- Built several full\-stack apps across the program — a learning\-management system, a banking app, and others — practicing REST APIs, auth, and responsive UI \- Strengthened core engineering fundamentals: JavaScript, API integration, and deploying working apps to production hosting

Web Development Intern at Tekzow, Jan 2025 \- Mar 2025 · 3 mos, Bengaluru, Developed full stack blogging platform with user authentication, role\-based access control, and Al\-powered content generation serving 100\+ test users\. Enhanced user experience by implementing dark mode, light mode toggle, command palette, and web accessibility features, improving accessibility score by 45%\. Built responsive user interfaces using React 18 and Tailwind CSS with smooth animations, achieving 95\+ Lighthouse performance score\. Integrated Progressive Web App capabilities enabling offline functionality and reducing load time by 35%\. Collaborated with 3\-member development team using Git version control and code reviews, reducing bug resolution time by 30%\. Participated in Agile development methodologies including sprint planning and daily standups\.

## Education

Bachelor of Engineering \- BE, Computer science engineering, Rajarajeswari College of Engineering\(RRCE\)

## Public links

- github: https://github\.com/HarshithNayakaL
- harshith\-nayaka\-l\-portfolio\.vercel\.app: https://harshith\-nayaka\-l\-portfolio\.vercel\.app/

[View the original profile](https://tal.club/meet/harshith-nayaka-l)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/harshith-nayaka-l#person",
      "name": "Harshith Nayaka L",
      "url": "https://tal.club/meet/harshith-nayaka-l",
      "image": "https://tal.club/meet/harshith-nayaka-l/profile-photo.jpg",
      "description": "Meet Harshith Nayaka L, AI Engineer, Full-Stack — LLM agents, RAG pipelines, multi-modal systems, production-grade output reliability | AI Workflow Engineer @ DemandNXT | Bengaluru based in Bengaluru. AI stack includes aistudio, anthropic apis, and claude code. Explore public work links including g...",
      "jobTitle": "AI Engineer",
      "worksFor": {
        "@type": "Organization",
        "name": "DemandNXT"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "AI Engineer"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bengaluru"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Rajarajeswari College of Engineering(RRCE)"
        }
      ],
      "knowsAbout": [
        "aistudio",
        "anthropic apis",
        "claude code",
        "claude",
        "codex",
        "openai"
      ],
      "sameAs": [
        "https://github.com/HarshithNayakaL",
        "https://harshith-nayaka-l-portfolio.vercel.app/"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/harshith-nayaka-l#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/harshith-nayaka-l#webpage",
      "url": "https://tal.club/meet/harshith-nayaka-l",
      "name": "Harshith Nayaka L on Tal",
      "description": "Meet Harshith Nayaka L, AI Engineer, Full-Stack — LLM agents, RAG pipelines, multi-modal systems, production-grade output reliability | AI Workflow Engineer @ DemandNXT | Bengaluru based in Bengaluru. AI stack includes aistudio, anthropic apis, and claude code. Explore public work links including g...",
      "dateModified": "2026-10-16T14:40:17.740Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/harshith-nayaka-l#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/harshith-nayaka-l/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/harshith-nayaka-l/profile-photo.jpg",
        "caption": "Harshith Nayaka L profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```