An engineer who ships scalable systems and production AI — end to end.
Senior Full-Stack & AI Systems Architect · 10+ years · Open to full-time (Remote worldwide / Relocation) · Available immediately.
Delivered $1M/year in cloud savings and −30% latency by re-architecting legacy healthcare & aviation monoliths with DDD.
Shipped a grounded, guarded RAG pipeline (OpenAI + pgvector) on WhatsApp for ZoetiCoach AI with an Approval Queue & Trust & Audit Log — sub-second, not a demo.
HIPAA-grade security (OAuth2/OIDC), 0-defect deploys with Pest/Vitest, and led a squad of 7.
*I read every message personally and reply within one business day.*
Snapshot
A scannable fact bar for recruiting teams| Status | Actively interviewing — open to full-time |
| Target role | Senior / Staff / Lead Full-Stack Engineer · Architect |
| Availability | Immediate / Flexible start |
| Location | India · Remote-first / Open to relocation for top-tier roles |
| Time zone | IST (UTC+5:30) · 4–6 hrs daily dedicated overlap with US (EST/PST) & full UK/EU overlap |
| Work authorization | Authorized to work in India · Open to Remote Worldwide & Visa Sponsorship/Relocation |
| Comp | Open to discuss based on role scope and impact |
What I'm looking for
A senior/staff role where ownership matters — taking systems from architecture through production, mentoring engineers, and shipping measurable outcomes. Strongest in fast-moving product teams modernizing complex domains (Healthcare, Aviation, Logistics, SaaS).
Why I'm worth a conversation
Cloud-Infrastructure Savings
Through architectural optimization, query tuning, and cache-aside layers at Infosys.
Clinical Latency Reduction
Clinical-trial monitoring latency via Redis Queues + real-time sync ($360K saved in 2024).
Cycle Time Reduction
Order-to-dispatch cycle time improvement through knitwear ERP legacy modernization.
User Efficiency Gain
Increase in user workflow efficiency modernizing legacy healthcare monoliths with DDD.
Engineering Leadership
Led technical squad; enforced Pest/Vitest automated testing standards for 0-defect deployments.
AI Engineering — production systems, not demos
I build LLM features that survive real users — grounded in retrieval, guarded against failure, and shipped inside products people already use.
RAG & Vector Search
Production Retrieval-Augmented Generation with pgvector embeddings (HNSW cosine search) and semantic re-ranking; grounded answers over habit evidence.
LLM Integration & Human Approval Queue
OpenAI APIs wired to an asynchronous Approval Queue — keeping coaches in the loop and preventing direct unreviewed LLM output to clients on WhatsApp.
AI Safety & Trust Audit Log
Multi-tier prompt-injection defense, hallucination evaluation, and full telemetry logging (tokens, latency, confidence score) for every AI event.
Agentic Automation
Autonomous accountability and intake workflows that turn loose chat threads into verified event-sourced ledger records.
How I work
Domain-Driven Design
Legacy monoliths → decoupled modular monoliths with clean bounded contexts and domain events.
Production AI & Vector RAG
Grounded RAG pipelines (OpenAI + pgvector) running live on WhatsApp with human-in-the-loop Approval Queue and full Trust & Audit logging.
Performance & Resilience
Event-driven architecture, Redis-first patterns, zero-downtime Strangler migrations, and Sentry observability.
Team Leadership & Standards
Mentoring engineers, fostering an RFC/code-review culture, and enforcing automated test discipline.
Core stack
Proof & Systems in Production
References & Verified Recommendations
LinkedIn RecommendationsFull recommendations and endorsements available directly on LinkedIn (Ashish Gupta) — alongside verified feedback from engineering leaders at Infosys, ZoetiCoach AI, and healthcare platforms. Happy to arrange direct reference calls upon request.
Let's talk about the role.
The fastest way to reach me is email or LinkedIn — I reply within a business day.