JG
NowSenior AI Engineer at Javis Technologies

I build AI agents that remember what you meant.

Hi, I'm Jatin, a Senior AI Engineer at Javis. I build AI agents that can follow long conversations, keep track of what's been agreed, and make sound decisions in real business workflows.

Bangalore, INB.Tech CSE · IIT Delhi
Jatin Goyal

Jatin Goyal

Senior AI Engineer · Javis

01 / Now · Javis Technologies

Building AI agents for real business work

Since November 2025 I've been a Senior AI Engineer at Javis in Bangalore, working on the core of its agentic AI platform.

Memory for AI agents

Helping agents remember what a conversation is about: what was asked, what was said and what was agreed. Follow-ups like "the second one" just work, and agents can pick up where they left off.

Conversations across chat and email

Extending that memory to long email threads with several people, so an agent keeps track of who said what and who committed to what.

Agents that make reliable decisions

Building blocks that let agents negotiate and schedule in business workflows, making consistent, explainable decisions rather than leaving everything to a language model.

Earlier at Javis

Voice AI agent

Built a real-time voice agent that answered live customer calls in English and Hindi, along with the phone system that connects callers to it.

02 / Experience

Where I've worked

  1. Nov 2025 — Present

    Senior AI Engineer

    Javis Technologies · Bangalore, IN

    Building AI agents that remember conversations and make reliable decisions in business workflows. Earlier, a real-time voice agent for customer calls.

    ↑ See the work above
  2. Jul 2022 — Nov 2025

    Software Engineer → Senior Software Engineer

    Enphase Energy · Bangalore, IN

    • Designed and built a Translation Management System that centrally manages i18n across a suite of microservices and microfrontends and delivers translations to services dynamically. PMs could update app text directly, with no code changes.
    • Improved search on the core Data entity by 95% with a denormalized Elasticsearch index that resolved data fragmentation across RDS and MongoDB, kept in sync in real time with Kafka and Logstash.
    Ruby on RailsElasticsearchKafkaLogstashMongoDBRDS
  3. Jul 2018 — Apr 2022

    B.Tech, Computer Science & Engineering

    Indian Institute of Technology, Delhi · New Delhi, IN

03 / Side projects

Things I built on my own

Screenshot of the ShortsKing web app
Solo startup · shut down

ShortsKing

archive.org

A fully automated platform for making short videos from minimal input: a title, a visual style and a voice. It organically attracted over 1,000 real users.

I ran the whole lifecycle on my own: design, frontend and backend, deployment, integrations and marketing. Content generation used open-source LLMs and ffmpeg.

Ruby on RailsOpen-source LLMsffmpeg

04 / Toolbox

What I work with

Agents & LLM systems

AI agentsLLM tool-callinggRPC / ProtobufDynamoDB

Voice & real-time

LiveKitWebRTCSIPFreeSWITCHOpenSIPS

Languages

PythonRubyJavaJavaScriptC++SQLKQLHTMLCSS

Frameworks & tools

FastAPIRuby on RailsSpring BootAirflowKafkaRedisDockerELK Stack

AWS

S3SQSMSKCloudWatchCloudFrontECR

Cloudflare

DNSWorkers AIR2CDN

05 / Contact

Let's talk.

Want to talk about AI agents or anything else? Get in touch.