Skip to content

Bengaluru · Open to AI engineering roles

I build AI systems that survive contact with reality.

AI engineer with full-stack depth, building production applications from 0 → 1—from autonomous agents and real-time voice to multimodal retrieval, APIs, infrastructure, and product interfaces.

Explore selected work ↘Let’s talk
Scroll to inspect the systems01 / 05

01 · Selected systems

Work with depth,
not demo-day polish.

Five products across security, sales, education, crisis intelligence, and generative media. Each one owns the hard parts behind the interface.

01

Autonomous cybersecurity

BountyKai CLI

A local, scope-enforced security assessment runtime coordinating specialist agents across web and Android, independently re-proving findings, self-healing broken tooling, and leaving auditable evidence on disk.

  • TypeScript
  • Python
  • Go
  • Multi-agent systems
↗02

Real-time voice + RAG

Marktiz Sales AI

An AI sales agent that joins live meetings, listens and speaks in real time, retrieves grounded product knowledge, then writes transcripts, summaries, follow-ups, and analytics to a dashboard.

  • React
  • Node.js
  • pgvector
  • Streaming speech
↗03

Multilingual voice teacher

GuruFlow AI

A voice-first teacher for Tamil, Hindi, and English. Learners interrupt with a doubt while deterministic agents rewrite the visual lesson, preserve memory, and resume the explanation.

  • Next.js
  • TypeScript
  • Voice AI
  • Structured agents
↗04

Multimodal crisis intelligence

Aegis

A command-center prototype turning video, images, radio, and field reports into cited situational intelligence with modality-aware retrieval, planner/critic agents, and provenance logs.

  • Python
  • Qdrant
  • Multimodal RAG
  • Agentic planning
↗05

Generative media pipeline

AI Character Studio

A fixed-identity production system spanning face discovery, LoRA training, ComfyUI orchestration, image and video generation, voice, talking avatars, and a full-stack control surface.

  • LoRA
  • ComfyUI
  • FastAPI
  • React
↗

02 · Operating mode

Think in systems.
Ship like an owner.

I’m an AI engineer who can work across the full stack and turn ambiguous ideas into dependable production applications.

01

End-to-end ownership

Architecture, interfaces, APIs, data, deployment, debugging, and iteration. I stay with the system until it works.

02

AI beyond the wrapper

Agent orchestration, RAG, streaming voice, multimodal pipelines, structured outputs, evals, and human verification.

03

Fast, without being fragile

Rapid execution backed by explicit state, scope controls, observability, critic gates, and failure recovery.

03 · Technical range

Across the stack.
Deep where it matters.

A working toolkit shaped by shipping complete AI products - from model orchestration and streaming infrastructure to the interface in a user’s hand.

01

Languages

  • Python
  • Go
  • Rust
  • TypeScript
  • JavaScript
  • SQL
02

Artificial intelligence

  • LLMs
  • Generative AI
  • RAG
  • Multi-agent systems
  • Voice AI
  • Multimodal AI
  • Embeddings
  • Vector search
  • LoRA
03

Backend & data

  • FastAPI
  • Node.js
  • Express.js
  • REST APIs
  • PostgreSQL
  • Qdrant
  • pgvector
04

Frontend & mobile

  • React
  • Next.js
  • Flutter
  • Tailwind CSS
05

Infrastructure

  • AWS
  • Azure
  • Docker
  • Vercel
  • Git
  • GitHub
  • Tauri
06

Engineering

  • Agent orchestration
  • Tool calling
  • Streaming systems
  • Structured outputs
  • API design
  • Debugging

Current focus Production agent systems · Real-time voice · Multimodal retrieval

04 · Experience

A track record of
turning zero into shipped.

BountyKai

Founder

Founded and built an autonomous AI cybersecurity platform combining AI agents, security tooling, backend services, and a desktop product. Own architecture and implementation across agent workflows, browser automation, security infrastructure, reliability, and product iteration.

KAT Enterprise LLC

System Analyst / Full-Stack Developer

Built and maintained production applications across backend services, frontend interfaces, APIs, databases, and deployment workflows. Independently owned features from ambiguous requirements through debugging and release.

TrendsToTravel

Founder

Built an LLM-powered travel discovery product focused on AI-assisted research and personalized discovery, leading it from problem definition and architecture through deployment and user-facing iteration.

TSOLE

Co-Founder

Co-founded and operated a SaaS services business serving 15+ clients, owning product delivery, technical execution, customer requirements, prioritization, and iteration.

05 · Contact

Hiring an AI engineer
who can ship 0 → 1?

I’m open to AI engineering and full-stack AI roles where I can own production applications end to end—from architecture through deployment.