Architect AI

Role Purpose:

We are looking for a sharp, self-driven AI-Enabled Architect who doesn't just use AI as a tool — but thinks, builds, and ships with AI at the core. This is not a traditional engineering role. You will operate at the intersection of deep technical execution and autonomous product thinking, leading end-to-end delivery with a lean team of AI agents as your collaborators.

You'll be building in the fintech/payments domain, which means your code handles real money, real compliance, and real users — speed and correctness are both non-negotiable.

If you've replaced standups with agent orchestration, and your shipping velocity is measured in ideas-per-week rather than tickets-per-sprint — we want to talk.

The responsibilities we entrust you

• Own the full build cycle — from research and architecture to deployment and iteration — with minimal human dependency.
• Design and ship RAG pipelines and autonomous agent systems (LangGraph, LangChain) capable of real-world reasoning, retrieval, and decision-making.
• Leverage AI coding tools (Cursor, Copilot, Claude, etc.) to write, review, and refactor production-grade code at speed.
• Build async Python services with FastAPI, Pydantic, SQLAlchemy 2.x, and Alembic — designing for correctness, idempotency, and scale.
• Implement secure, multi-tenant platforms with OAuth2, JWT, RBAC, secret management, and OWASP-compliant practices.
• Integrate payment flows — webhooks, refunds, settlements, mandates, and UPI AutoPay — with full awareness of PCI requirements.
• Author TypeScript SDKs and frontend features using React + Vite with strict typing, monorepo tooling, and backwards-compatible versioning.
• Ship and operate on Kubernetes — writing Helm charts, managing Docker Compose for local dev, and owning CI/CD via GitHub Actions.
• Instrument everything — OpenTelemetry, structlog, Prometheus/Grafana — because observability is not optional.
• Wear the product hat — translate research outcomes into features, write specs, and make trade-off decisions when required.
• Conduct deep research using AI tools to evaluate architectures, benchmark models, and validate technical decisions independently.
• Build a new connector + tool registration, ship an SDK release, implement an API endpoint with RBAC + migration, extend Lang Graph agent nodes, write integration tests, triage prod issues.
• Design the connector framework, agent orchestration, auth/tenancy model, payment flow (mandate → pay → settle), observability, Helm/infra. Drives code review and threat modelling..

What matters in this role:

• Autonomous by default — you don't wait for answers; you find them.
• Research-first thinker — comfortable diving into papers, docs, and benchmarks to make informed decisions.
• Product-minded engineer — you understand what you're building and why, not just how.
• High agency — can define the problem, architect a solution, and ship it, end-to-end.
• Comfortable with ambiguity — early-stage thinking doesn't scare you; it energizes you.
• Security-conscious — treats auth, secrets, and data handling as first-class concerns, not afterthoughts.

Experience & Portfolio:

Backend & Platform (Must-Have)

• Python 3.11+ — async/await patterns, httpx, pydantic, FastAPI, SQLAlchemy 2.x, Alembic
• LLM / Agentic systems — LangGraph, LangChain, tool-calling, prompt design, OpenAI/Gemini APIs, RAG pipelines, evaluation frameworks
• Databases — PostgreSQL (JSONB, migrations), Redis,vector DB etc
• Auth & security — OAuth2, JWT, RBAC, multi-tenancy, secret management, OWASP Top 10, TLS/truststore configuration
• Testing — pytest, Vitest, contract tests, load/performance testing

Frontend & SDK (Must-Have)

• TypeScript (strict mode), Node 20+, monorepo/package authoring, tsup/tsc build pipelines
• React + Vite, modern state management patterns
• SDK design — versioning strategy, backwards compatibility, typed client generation

DevOps / SRE (Must-Have)

• Observability — OpenTelemetry instrumentation, structlog, Prometheus/Grafana dashboards
• CI/CD — GitHub Actions, release automation, environment promotion pipelines

Nice to Have

• Kotlin + Ktor + Gradle (for cross-platform subscription services)
• MCP (Model Context Protocol) server/client implementation
• RPA / browser automation with Playwright
• Compliance familiarity — SOC2, RBI/PA-PG guidelines for Indian payments
• Docker, Docker Compose, Helm, Kubernetes — comfortable writing and maintaining manifests end-to-end

Technical skills & efficiency:

• 10–15 years of overall software engineering experience
• 2–3+ years of hands-on AI/ML engineering in production environments
• Proven ability to ship with AI-first workflows — using LLMs and agents to code, test, and deploy with minimal human intervention
• Strong research instinct — can independently benchmark, evaluate, and decide on architectures using AI tools
• Experience building event-driven, scalable distributed systems
• Comfortable working in a fintech environment where reliability, security, and auditability matter
• Product-minded — can translate research into specs and own features end-to-end

Communication & Ownership:

1. Strong communication, interpersonal, and conflict resolution skills
2. Able to work both independently and as part of a team

Location: Sector 62, Noida

This is 5 days work from office role (No remote or hybrid)

Things you should be comfortable with:

1. Working from office: 5 days a week
2. Pushing the boundaries: Have a big idea See something that you feel we should do but haven't done We will hustle hard to make it happen. We encourage out of the box thinking, and if you bring that with you, we will make sure you get a bag that fits all the energy you bring along.

What we value in our people:

1. You take the shot: You decide fast and deliver right.
2. You are the CEO of what you do: You show ownership and make things happen.
3. You sing your work like an artist: You seek to learn and take pride in the work you do.