RKRahul Khanna D

Generative AI Engineer

shipping LLM systemsthat survive production.

Seven years of robots, factory vision and embedded devices — and now LLM apps, RAG pipelines, AI agents and MCP integrations, built with the same discipline.

Start a conversationView mission files ↓

Scan complete · Target: production-grade GenAI · Acquired

Briefing Console

Initialising

RK

Linking profile · loading mission files · arming guardrails

Try
›

01Diagnostics

System check: all readings nominal.

7+

years shipping production AI

55+

innovation awards

250+

open-source projects

100K+

learners on Electroverse.AI

2

patents

50+

certifications

Deployed for

  • ITC
  • Godrej & Boyce
  • Ashok Leyland
  • BHEL
  • HCL
  • Lucas TVS
  • Thejo Australia
  • Agnikul
  • IIT Madras Research Park

02Mark I – IV

Built in iterations, like any good suit.

Every suit was built on the one before it. So was this career: each stage carried its lessons into the next.

  1. Mark I

    2018 — 2019

    Research

    IIT Madras Research Park & HTIC. An AI model to predict LFIA2 outcomes for research-grade medical devices, and IMU-based motion tracking with sensor fusion.

    Proven in the lab

  2. Mark II

    2019 — today

    Edge AI & Robotics

    Autonomous mobile robots from 100 kg to 2000 kg payload, AI visual inspection, pallet tracking — deployed for Tier-1 manufacturing, government and SpaceTech clients.

    Deployed in the field

  3. Mark III

    Now

    Generative AI

    LLM applications, RAG pipelines, AI agents, tool calling and MCP servers — with evaluation and guardrails built in, down to quantized LLMs running on-device.

    In production

  4. Mark IV

    BITS Pilani

    Quantum ML

    Variational quantum circuits, quantum feature maps and hybrid quantum-classical models — the next layer on top of the GenAI stack.

    M.Tech specialization

03Mission files

Ten missions. One file each.

Pick a file. Each one shows what the system does, what it was built with, and where to see it.

Mission file 01 / 10 · GenAI

Live

Four Takes

System schematic

  1. 1Short video in
  2. 2Gemma grounds captions to the footage
  3. 3Four persona-styled captions
  4. 4LLM-as-judge scores accuracy and tone
  5. 5Scored takes ship
LiveOpen live demo ↗

Objective

A live multimodal LLM app that turns short videos into four persona-styled captions in one pass. Gemma grounds the captions to the actual footage; an LLM-as-judge stage scores every take on accuracy and tone before it ships.

Stack

  • Multimodal LLM
  • Gemma
  • LLM-as-Judge
  • Next.js
Open live demo ↗

04Service record

Seven years of AI that had to work the first time.

  1. Apr 2021 — Present

    Sakthi Robotics

    Co-Founder & Technical Director, Products

    • ›Built LLM- and NLP-based human-robot interaction systems, including on-device quantized LLM inference for natural-language command understanding on AMR platforms.
    • ›Developed RAG-based knowledge retrieval over product documentation, onboarding guides, and deployment manuals.
    • ›Led multiple AI deployments in production robotics — data pipelines, model integration, edge inference under latency and reliability constraints.
    • ›Built AI-based pallet tracking and visual-intelligence auto-part inspection (SPM) systems.
    • Government innovation grant
    • Multi-product AMR & pallet-shuttle roadmap
    • Finalist, Automate Startup Challenge 2024
    • Top 8% WTFund by Zerodha

  2. 2019 — Present

    Sakthi CADD

    CTO — Hardware, Edge AI & Product Strategy

    • ›Led tech roadmap and product execution for Tier-1 manufacturing, government, enterprise, and SpaceTech clients.
    • ›Introduced AI-assisted engineering workflows (Claude Code, Copilot) across dev teams.
    • ›Bridged research prototypes to deployment-ready ML systems.

  3. 2018 — 2019

    IIT Madras Research Park & HTIC

    Research Experience

    • ›Developed an AI model to predict LFIA2 outcomes — data pipelines, model experimentation, and validation for research-grade medical devices.
    • ›Built IMU-based joint monitoring and motion-tracking systems using sensor fusion.

05Arsenal

From the prompt to the PCB.

Evaluation and guardrails come standard. Hasta la vista, hallucinations.

LLMs for Generative AI

Building and shipping LLM applications, end to end.

13 skills

  • LLM application development
  • RAG architectures
  • AI agents
  • Prompt engineering
  • Function / tool calling
  • MCP servers
  • Claude API
  • OpenAI API
  • LangChain
  • LangGraph
  • Vector search & embeddings (FAISS, Chroma)
  • Evaluation & guardrails
  • On-device quantized LLM inference

06Training & commendations

Recognised by the names that build the hardware.

Role

ARM Ambassador

Role

Intel Edge AI Ambassador

Role

Arduino Certified Instructor

Role

Hackster.io Pro

Electroverse.AI

YouTube channel for robotics & AI educational content, reaching 100K+ learners.

Watch the channel ↗

Awards

40+ national and 15+ international — Intel, Microsoft, Avnet, Schneider Electric, Particle.

Grant

Government of India innovation grant, Ministry of Science & Technology.

2025 — 2027

BITS Pilani

M.Tech (Executive), AI & ML

Specialization: NLP & Information Retrieval · GPA 8.75

2016 — 2020

Saveetha Engineering College

B.E., Electronics & Communication

GPA 8.04

Udacity Nanodegrees

  • Intel Edge AI for IoT DevelopersScholarship — selected from 17,000 applicants
  • Intro to Self-Driving Cars
  • AWS Machine Learning Scholarship

Awards from: Intel · Microsoft · Avnet · Schneider Electric · Particle · ARM · Arduino · Hackster.io · Element14

07Open a channel

No fate but what we ship.

Building something with GenAI? I'm open to Generative AI engineering roles — remote, or relocating from Chennai.

Start a conversation

© 2026 Rahul Khanna D

No Skynet was trained in the making of this site.

I'll be back ↑