Skip to content
Open for select engagements

Work with me.

I am building InfiniTraq as Co-Founder & CTO of Griffin AI Tech. Alongside, I take on a small number of part-time engagements each year — the kind where a decade of data and AI systems experience compounds quickly into your team.

What I take on

Three lanes where I'm most useful.

The work I do best sits at the intersection of data systems, applied AI, and early-stage engineering leadership. The lanes below are the ones where I can move the needle fastest and where I genuinely enjoy the problem.

Data Platform & Lakehouse

Architect, audit, or scale your data stack.

  • Green-field lakehouse design (Iceberg / Hudi / Delta)
  • Spark / Trino / Hive / HBase tuning at PB scale
  • Airflow / Kafka / Maxwell / DeltaStreamer pipelines
  • AWS EMR, S3, Athena, Lambda, Kinesis cost & perf reviews
  • Migration from legacy warehouses to open table formats

Applied AI & Automation

Vision, language, and the glue between them.

  • Computer vision pipelines — detection, tracking, classification
  • LLM & RAG document intelligence (LangChain + OpenAI + vector stores)
  • Workflow automation with n8n, Supabase, and custom APIs
  • Edge AI on NVIDIA Jetson / Orin for real-time inference
  • MLOps: training, eval, registry, monitoring
  • Domain background: medical imaging, video safety, document intelligence

Fractional CTO & Advisory

For seed and Series-A teams that need senior leverage.

  • Architecture review & technical due diligence
  • Hiring help — JD design, interviewing, calibrating bar
  • Engineering coaching for first-time founders & tech leads
  • Sprint cadence, on-call & incident reviews
  • Investor / board technical narrative support

How I work

Quiet, async, written.

Async-first, decisions written down, one weekly working session. I optimise for your team's focus, not for billable hours. Each engagement starts with a one-page scope brief so we both know what success looks like before anything meaningful changes in your code.

01

Intro call

A short async exchange or a 30-min call to understand the problem, the stakes, and whether I am the right fit. No commitment.

02

Scope brief

I write up a one-page brief — outcomes, scope, timeline, working rhythm. You decide if it lands before anything else moves.

03

Engagement

Fixed-price project, weekly retainer, or advisory hours. Async-first, with one weekly working session and shared docs / Linear / Notion.

04

Handover

Documented, with a checklist your team can run from. I stay reachable for follow-ups for a fixed window after wrap-up.

Fit

A short list. Both ways.

The honest filter — what tends to go well, and what I steer clear of. If you are unsure which side you fall on, write to me anyway; I will tell you straight.

Good fit

  • Founders / CTOs needing a senior pair of eyes on data or AI architecture
  • Teams scaling past their first PB or first real production ML system
  • Companies inheriting a legacy data stack and weighing a rebuild
  • Seed / Series-A teams that want a fractional engineering partner, not a contractor

Not a fit

  • Full-time roles or anything resembling a long-tenure operator job
  • Generic feature-factory dev work with no architectural problem to solve
  • Equity-only or contingent-pay engagements
  • Anything that materially conflicts with my work at InfiniTraq

Receipts

Outcomes from past decade.

A few concrete results from the work behind the resume. Full case studies under Projects.

14 PB

Data lake at Shopee SG

Architected and operated a 14 PB lakehouse serving thousands of internal users across analytics and ML.

30K+

Pipelines at 99.95% reliability

Led ingestion for over 30,000 data pipelines at Shopee, holding 99.95%+ job-success SLOs through self-healing, reconciliation, and disciplined on-call.

<5 s

Edge AI alerting

Designed the InfiniTraq Edge AI pipeline that detects falls and inactivity on-device with sub-5-second response.

8

OSS libraries shipped

Maintained 8 Ruby gems and other OSS used in production by Rails teams.

In their words

A decade of working alongside good people.

From the founders, managers, and clients I've shipped with — paraphrased verbatim from LinkedIn and Upwork.

Ethiraj is great when learning new stuffs and implementing it in short span of time, he is keen to technology and surely a good assets that any team can have. It was nice working with Ethiraj on couple of projects together. All the best for future endeavours.
Sunil Prakash

Enterprise AI & Platform Architecture Leader · VP, Regulated Finance · Author, Agentic AI for Serious Engineers

via LinkedIn
Good learn ability and think fast. Work hard with good ownership. Good team mindset. It is my luck to work with you.
Zhaoqun Deng

Expert Engineer · Shopee

via LinkedIn
Ethiraj is an excellent communicator and gave us ideas on improving the data pipelines. Strong data engineering knowledge. Highly recommended.

Manish M.

Data Engineering Client · Upwork

via Upwork
Ethiraj is really a cool person to have in team. His passion towards technologies is really astonishing. He has got the right mix of knowledge, passion, adaptability and ownership. He has always followed high coding standards and contributes for open source as well.
Martin Devapitchai

Director & Co-Founder · Hash Agile Technologies

via LinkedIn

FAQ

Common questions.

Where are you based and what hours do you work?

Chennai, IST (UTC+5:30). I work async-first with one weekly working session in the client time-zone. I have worked extensively with teams in SG, EU, and US.

How do you price engagements?

I price based on the scope and shape of the work — fixed price for clearly defined projects, weekly retainer for ongoing engineering or advisory work. We discuss it on the intro call once I understand the problem.

Can you sign NDAs and DPAs?

Yes. I am happy to sign reasonable mutual NDAs before the scope brief, and DPAs as required by your jurisdiction.

Will my work conflict with InfiniTraq?

No. I only take engagements that are non-competing and that I can do well alongside InfiniTraq. If there is any conflict, I will name it on the intro call and walk away.

What does a typical engagement look like?

Common shapes: a 2–4 week architecture review, a 6–12 week build / migration project, or a multi-month advisory retainer of a few hours per week. We will pick whichever shape actually fits the problem.

Start a conversation

A short note is enough. Tell me what you're building.