Skip to content
View Nikhils-G's full-sized avatar
🪄
I try to commit everyday it was not happening .Even Though I still try !
🪄
I try to commit everyday it was not happening .Even Though I still try !

Highlights

  • Pro

Organizations

@Powersmybiz

Block or report Nikhils-G

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Nikhils-G/README.md
nikhil@ai-agents:~$ whoami — Nikhil Sukthe, AI Engineer, building agents that automate business across voice, WhatsApp, chat and email

I build AI agents that automate real business and get measured on outcomes — voice agents that talk on the phone (LiveKit + SIP, streaming STT-LLM-TTS), WhatsApp commerce agents that take a customer from catalog to Razorpay payment without leaving the chat, and the chat & email automation around them. Founding AI Engineer at Fika.ai. The metric is never the demo — it's collection rates, bookings, and latency. Hyderabad, India.

Open source

livekit/agents #4798 · fix(sarvam): prevent transcript loss after long agent responses

A race condition in the Sarvam STT plugin: agent responses longer than ~10 s triggered premature task cancellation while buffered audio was still being processed — transcripts were silently dropped and the agent went mute mid-call. Found it running live voice-agent operations; fixed it upstream.

Now

Building Multi-channel AI agents (voice · WhatsApp · chat · email) for healthcare, fintech and real-estate clients at Fika.ai
Running Live voice-agent ops — SIP trunks, STT/TTS provider A/B tests, per-call cost & latency telemetry
Exploring Speech-to-speech (Gemini Live), predictive & outbound dialers, MCP tool servers

Selected work

Project The number that matters
Voice WhatsApp Voice Bridge WhatsApp Desktop → WASAPI → LiveKit → Gemini Live: two-way audio at 300–800 ms RTT, zero SIP cost
Voice LiveKit SIP Manager Full-stack SIP ops dashboard — trunks, dispatch rules, IP whitelisting, dialer — zero CLI
Agents NBFC collections voice agent Production collections calls at Fika.ai — collection rate up 50%
Agents LangGraph + RAG assistant Sub-5 s end-to-end latency, 15% engagement uplift
Lab ai-experiments-lab Reproducible experiments across ML · DL · NLP · LLMs, one concept at a time

More on nikhilsukthe.vercel.app.

One runtime, every channel

Nikhil's stack drawn as a circuit board: Phone/SIP, WhatsApp and Chat/Email inputs flow into an Agent Runtime chip (LiveKit, WebRTC, SIP, STT and TTS, Gemini Live, Sarvam) wired to an LLM Core chip (LangGraph, MCP, RAG, LoRA/QLoRA), mounted on an infra bus of FastAPI/WebSockets, MongoDB/Redis/Postgres, Docker/Kubernetes, AWS/GCP Vertex, Langfuse, and vector databases

Live stats

Neofetch-style card with live GitHub stats for Nikhils-G

Recent activity


Website · LinkedIn · Email · ORCID

This profile updates itself — a daily GitHub Action runs one Python script that regenerates the stats card and activity feed. Last refreshed: 2026-08-25 (UTC).

profile refresh status

Pinned Loading

  1. ai-experiments-lab ai-experiments-lab Public

    AI Lab - which is Core Understanding & Practical Experiments in depth with each functionality.

    Jupyter Notebook 1

  2. SystemDesign-HighLevel SystemDesign-HighLevel Public

    Python 1