Senior Full-Stack AI Engineer

Experience 3.00 + years Salary Confidential (based on experience) Shift (GMT+05:30) Asia/Kolkata (IST) Opportunity Type Remote Placement Type Full time Permanent Position

**(*Note: This is a requirement for one of Uplers’ client - CellStrat)

**What do you need for this opportunity?

Must have skills required: Python, FastAPI, Agentic AI, multi-agent AI systems, react, Next.js, TypeScript

CellStrat is Looking for:** CellStrat is an AI startup with two sides to the business. The services arm builds end-to-end AI engineering solutions for US-based clients across industries. The product side is CellAssist, a healthcare AI platform with a voice-first approach — AI receptionists, automated patient triage, and admin operations powered by voice agents.

This role sits primarily on a US client project (80%), with a partial contribution to CellAssist (20%). As a senior engineer, you will own features end to end, shape architecture, review code, talk directly to clients, and be expected to figure things out without being handed a spec.

• What You’ll DoOwn features end to end — design, build, review, test, release, and monitor what you ship.
• Design and build backend systems in Python and FastAPI, from architecture to production.
• Architect SaaS platforms and reason about platform-level concerns: multi-tenancy, service boundaries, data models, async workflows, and infrastructure.
• Design and build agentic and multi-agent AI systems from first principles: agent orchestration, tool and MCP server integrations, memory and state handling, evals, and guardrails.
• Apply production LLM engineering — prompt engineering, context management, memory strategies, and prompt caching — to control latency, cost, and reliability.
• Review code across the team with a sharp eye for correctness, design, security, and maintainability, including AI-generated code.
• Contribute to frontend work in React, Next.js, or TypeScript when needed.
• Join US client calls (night IST) and do the work of understanding the domain, not just collecting feature tickets.
• Use AI-assisted development tooling (Cursor, Claude Code) to ship fast without cutting corners.
• Must-Have Requirements
• Python as primary backend language — FastAPI is the default framework; intermediate-to-strong fundamentals (OOP, decorators, generators, async); not a Node.js-first engineer who recently added Python.
• Production-grade agentic or multi-agent AI system — built and shipped at least one agentic system used by real users; can speak to user feedback, challenges, and production constraints.
• Software architecture ownership — has designed systems from scratch: HLD, LLD, service boundaries, data models, async workflows, multi-tenant patterns; can explain decisions and trade-offs in plain language.
• Trade-off reasoning — can justify technology choices over alternatives (e.g. why RAG vs. no RAG, why a particular vector DB or deployment method); does not default to stacking frameworks without rationale.
• Actively writing code — still hands-on daily; uses AI coding tools (Cursor, Claude Code) to accelerate, not replace, their own coding ability.
• Night IST availability — genuinely comfortable with regular evening/night US client calls (1–1.5 hrs; daily, alternate, or weekly depending on client); available for occasional weekend production support.
• LLM-layer engineering — prompt engineering, context management, memory management, and prompt caching in real production systems.
• React, Next.js, or TypeScript working knowledge — enough to make feature changes and debug UI issues; full frontend ownership not required.
• Good to HavePure Python agent implementation without LangChain or LlamaIndex — built agents from first principles.
• Active GitHub profile with recent Python/FastAPI projects (not predominantly Jupyter Notebooks).
• GCP experience beyond Cloud Run (GCS, IAM, Pub/Sub, Cloud Tasks).
• LLMOps familiarity — evals infrastructure, prompt versioning, AI observability in production.
• Experience leading design reviews or mentoring through code review.
• Custom AI-assisted development workflow built or customised beyond IDE defaults.
• Prior US client-facing engineering experience.
• Interview ProcessRound 1: Proctored live coding — HM joins for 5 minutes to explain; candidate fixes missing Python/FastAPI code in a provided zip project within 20 minutes (recorded).
• Take-ho…