AI Systems Engineer

I build AI systems that survive contact with reality.

Video intelligence, analytics platforms, study copilots. I build the product, backend, and system design, then deal with the part where reality starts filing bug reports.

Years

5+

shipping production-grade software

Shipped

8+ apps

across AI, SaaS, fintech, and IoT

APIs

130+

across NeuralEye and Tracklink

Designation

SDE-2

current role and scope

Selected Systems

The projects I use when someone asks what I actually build.

AI systems, analytics platforms, and just enough operational caution to keep the interesting parts alive.

AI Video Analytics

flagship

NeuralEye

A production-grade video intelligence platform for zero-shot brand and logo detection at scale.

01

Built as a multi-tenant AI system that combines zero-shot proposals, reference normalization, multi-modal verification, distributed inference, and observability to process long-form video without leaning on constant retraining.

System snapshot

APIs

50+

across ingestion, detection, analytics, and reporting

Efficiency

2x

through layered inference, batching, and queueing

Architecture

Multi-modal

zero-shot proposals, reference prep, and verification

FastAPI

React

TypeScript

PostgreSQL

Redis

Celery

Open dossier

Zero-shot detection, FAISS matching, async orchestration, and a dashboard for teams that prefer their pipelines observable.

Analytics SaaS

02

Tracklink

shipped

A multi-tenant analytics platform for tracking, segmentation, and link intelligence.

Designed as a scalable SaaS platform with Go, React, Redis-backed analytics, RBAC, and high-volume ingestion pipelines for real-time reporting.

Event flow

captureaggregatereport

Go

Fiber

React

TypeScript

Open dossier

AI Learning Product

03

CoStudy

in progress

An AI study copilot that helps students and professionals learn with structure, context, and multi-model support.

Designed to turn AI from a chat tool into a real study workflow, combining guided learning, reasoning assistance, and model orchestration for better academic outcomes.

Study flow

React

Next.js

TypeScript

AI orchestration

Open dossier

Build Approach

How I build once the nice, tidy version stops being true.

Product clarity, engineering depth, and enough structure that the next six months do not turn into archaeology.

AI Systems

01

I build AI systems that still make sense once they meet users, latency, and product constraints.

Signature Moves

Multi-stage inferenceModel routingWorkflow-first AI UX

Tool Deck

FastAPI

Celery

FS

FAISS

React

TypeScript

/systems/neuraleye /systems/costudy

Backend & APIs

02

I build service layers that stay readable after the fourth feature request and the second "we should rewrite this" meeting.

Signature Moves

RBACREST APIsMulti-tenant architecture

Tool Deck

Node.js

Go

FastAPI

PostgreSQL

Redis

/systems/tracklink /systems/neuraleye

Frontend Product Systems

03

I care about product clarity as much as architecture, so the UI has to turn complexity into decisions instead of decoration.

Signature Moves

SSRDesign systemsDashboard UX

Tool Deck

React

Next.js

ZU

Zustand

RQ

React Query

TypeScript

/systems/costudy /systems/neuraleye /systems/tracklink

Real-Time & Observability

04

I think about what happens after release: retries, metrics, logs, alerts, and enough telemetry to keep ambitious systems honest.

Signature Moves

MonitoringEvent streamsFault tolerance

Tool Deck

Prometheus

Grafana

Loki

Redis

WS

WebSockets

/systems/neuraleye /systems/tracklink

Path

Career arc, minus the LinkedIn theater.

2022 - now

01

Software Engineer

Procedure Technologies

Moved into full-stack ownership across AI, analytics, SaaS, architecture, mentoring, and the part where complexity starts charging interest.

2021 - 2022

02

Junior Software Engineer

PrismScale

Got fast at shipping interfaces, product flows, and frontend systems that had to hold up outside tutorial conditions.

2020 - 2021

03

Intern

BrandSatvam

Learned the useful fundamentals early: React, FastAPI, data models, auth flows, and the habit of cleaning systems instead of decorating them.

Team Impact

I like leaving both codebases and people less confused than I found them.

Mentoring, documentation, reviews, and the quiet work that keeps teams fast without turning them frantic.

Mentoring Engineers

I help newer engineers get productive faster through reviews, pairing, and clearer technical framing.

Less hesitation, better reviews, and a team that ramps without drama.

Owning Systems End to End

I move from requirements to implementation, testing, deployment, and iteration without dropping the product thread halfway through.

Delivery stays cohesive across frontend, backend, and infrastructure.

Creating Reusable Engineering Context

I document decisions, structure codebases for reuse, and try not to leave future teammates a detective story.

Onboarding gets easier, and the codebase keeps more of its dignity over time.

Contact

If you're building something intelligent, useful, and slightly unruly, we should probably talk.

I'm most interested in AI-first products, analytics systems, full-stack platforms, and teams that like ambition but still leave room for taste.

Say hello Find me on LinkedIn

I build AI systems that survive contact with reality.

The projects I use when someone asks what I actually build.

NeuralEye

Tracklink

CoStudy

How I build once the nice, tidy version stops being true.

AI Systems

Backend & APIs

Frontend Product Systems

Real-Time & Observability

In regular rotation

Production muscle

Current edge

Career arc, minus the LinkedIn theater.

Software Engineer

Junior Software Engineer

Intern

If you're building something intelligent, useful, and slightly unruly, we should probably talk.