AI-assisted communication
Tools like ReplySeq, an AI WhatsApp Business reply assistant currently in development, that drafts replies from a business's own product information while keeping a human in control of what actually gets sent.
ReplySeq
Engineering AI systems. Building AI products.
I'm an AI and backend engineer who takes systems from a rough idea to something running in production — LLM pipelines, RAG, and automation that hold up under real traffic. Alongside client work, I build independent products under SeqOrbit.
Currently building production LLM systems at StratSync (Singapore) · Founder of SeqOrbit
🏆 ₹4L Hackathon ChampionFeatured in Dainik BhaskarPowering a Govt Smart-City PlatformLive in a Singapore SaaS
Backend & APIs
Production services and architecture
LLMs & RAG
AI systems built for real usage
Product Engineering
From research through deployment
Automation
Workflows and tool integrations
Government collaboration proof
Available for projects
AI Systems Engineer
StratSync Pte Ltd, Singapore
Building production LLM systems, RAG pipelines & backend infrastructure for real users - not demos.
Founder of SeqOrbit
AI Product Studio
Building independent AI products end-to-end — product direction, engineering, and shipping.
GovTech AI Builder
ISCDL - Smart City Initiative
Built AI + GIS tooling for illegal-construction detection at government scale.
About
I take ideas all the way
through production.
I'm a software and AI engineer who likes taking an idea all the way through — research, architecture, build, deployment, and then the part most people skip: iterating on it once it's actually being used.
Most of my work lives at the intersection of AI and backend systems: LLM integrations, RAG pipelines, vector databases, and the APIs and infrastructure that keep them running reliably, not just in a demo. I've applied this both in client work — building production systems at StratSync — and in government-scale deployment, building AI and GIS tooling for a Smart City initiative.
Outside client work, I run SeqOrbit, where I build and ship independent AI products. I'd rather own a system end-to-end — from the model layer down to deployment — than hand off pieces of it and lose the thread.
What I Do
What I Do
Everything below has shipped somewhere real — a Singapore SaaS, a government Smart City platform, or a product I built myself under SeqOrbit.
$ whoami --studio
Founder of SeqOrbit.
SeqOrbit is an AI product studio I started to build practical, AI-powered tools — the kind that solve a real workflow problem rather than add another dashboard to check. It's early-stage: no outside roadmap, no client briefs. Every product starts as a problem I've run into myself, and it only ships once it actually solves that problem.
The idea behind it — build in sequence, scale in orbit — is about focus: build one thing well before starting the next, rather than spreading across five half-finished ideas at once.
As founder, I'm doing the building myself: product direction, engineering, and shipping. Current focus areas include AI-assisted communication tools and workflow automation for professionals and small businesses.
Under SeqOrbit
What I'm Building
Current focus areas under SeqOrbit — some shipped, some still in progress.
Workflow automation
Connecting the tools professionals already use so fewer things need manual handling.
Browser-native tools
Lightweight extensions that sit inside an existing workflow instead of asking someone to switch to a new app.
Meeting & communication productivity
Early exploration through MeetSeq; the direction is still being shaped.
MeetSeqThis section will keep growing as products move from idea to shipped.
Career
Experience
Professional work and independent building, with government-scale execution as supporting proof.
AI Systems Engineer
CurrentStratSync Pte Ltd · Singapore
ActiveProfessional WorkRemote
Building production LLM systems and RAG pipelines, along with the backend infrastructure they run on. Focused on systems that handle real users, not proof-of-concepts.
Founder of SeqOrbit
OngoingSeqOrbit · Independent
ActiveProduct StudioFounder-led
Building independent AI products end-to-end — product direction, engineering, and shipping.
GovTech AI Builder
2025Indore Smart City Development Ltd. (ISCDL) · Indore, India
Government CollaborationPiloted
Built AI and GIS tooling for illegal-construction detection at government scale, presented to IAS-level officials.
Key Achievements
Capabilities
Technical Expertise
AI & LLMs
LLM integration, RAG pipelines, vector databases, prompt/context engineering, agentic workflows
Backend
API development, backend architecture, service design
Frontend
React, Next.js
Automation & Integrations
n8n, Microsoft Teams, Slack, browser/Chrome extensions
Cloud & DevOps
Docker, GCP, CI/CD, GitHub Actions, monitoring & testing
Databases
Vector databases, relational databases
Proof of execution
Selected Work
Smart City Construction Monitoring
AI + GIS system for detecting illegal construction at city scale.
- Problem it solves
- Manual monitoring across a city is slow and inconsistent.
- What I built
- Built the AI/GIS detection pipeline and dashboard, presented to district officials.
- Status
- Piloted with ISCDL (LOI issued)
- Technology
- AI/GIS, geospatial data pipelines
SeqOrbit
ReplySeq
An AI WhatsApp Business reply assistant for small businesses.
- Problem it solves
- Small businesses lose time and consistency replying to customers manually.
- What I built
- Building the reply-drafting engine using the business's own product/knowledge data.
- Status
- In development
- Technology
- LLM integration, WhatsApp Business API
Recognition & Impact
Smart City & GovTech Collaboration
1st place at the Indore Tech Hackathon 2025 led to an official collaboration with Indore Smart City Development Limited (ISCDL) — building AI + GIS systems for urban monitoring, backed by an issued LOI.
- Prize Won
- ₹4 Lakhs
- Rank
- 1st Position
- Status
- LOI Issued
Process
My Approach to Building
I start with a real problem — usually one I've hit myself — before writing any code. I'd rather ship a small, usable version fast and see how it's actually used than spend weeks building something no one's touched yet. AI goes in where it removes real work, not because it's expected. Then I iterate based on what usage actually shows, not what I assumed going in.
Collaboration
Why Work With Me
How I approach ownership, execution, and communication.
01
I ship, not just build.
Everything I work on has gone live and handled real usage — not code that stalls after handoff.
02
I own the full stack.
AI layer, backend, APIs, infrastructure — I handle these myself, which means fewer handoffs and faster iteration.
03
I've built under real pressure.
Presenting to IAS-level officials and deploying for a government initiative isn't a comfortable environment — it sharpens how I work under constraints.
04
I think in systems.
Before writing code, I think about scale and failure modes, so what I build is designed to grow, not just demo well.
05
I stay current, deliberately.
I track what's actually useful in AI tooling — RAG patterns, vector DBs, new model APIs — and adopt what matters, not what's trending.
06
Clear communication.
You know what's in progress, what's blocked, and what's next, without having to chase me.