Backend & Integrations Engineer (iris)

About iris & Affinity Answers

Affinity Answers is a self-funded, profitable global company that helps 400+ advertisers, agencies and publishers across the US, EU and APAC understand and target human behaviour.

iris, incubated by Affinity Answers, is a fully managed patient-growth platform for independent clinics in India. It gets clinics found on Google, Maps and AI answers, chosen through their Google Business Profile and reviews, and booked through Iris Concierge, which engages, qualifies and books leads inside the WhatsApp Business app clinics already use. Clinics go live in 7 days.
Almost everything iris does depends on reliable integrations with Google and Meta. That is the core of this role.

What you'll build and own

• Google integrations - Google Business Profile (posts, reviews and replies, insights, locations), Google Search Console (performance data, URL inspection, sitemaps) and Google Ads, built so each clinic connects its own account in a few clicks.

• Meta integrations - WhatsApp Business Platform (Cloud API, Embedded Signup and coexistence with the WhatsApp Business app), Instagram Graph API (publishing, insights, messaging), Meta Marketing API (campaigns, lead ads, insights) and Conversions API.

• OAuth done right - Authorization-code flows, scopes, consent screens, token storage and encryption, refresh and rotation, revocation and re-auth, multi-tenant account linking per clinic.

• App verification and reviews - Own the paperwork and the engineering for Meta App Review and Business Verification (Tech Provider where needed), Google OAuth app verification for sensitive scopes, GBP API access requests and the Google Ads developer token, through to production approval.

• AI agents and RAG - Build the AI inside iris: the Iris Concierge agent that qualifies leads, follows up and books consultations through tool calls; RAG over each clinic's treatments, pricing, FAQs and doctor profiles so answers stay accurate per clinic; and LLM pipelines that draft Google posts, blogs and audit insights. Own prompts, evals, guardrails against unsafe medical claims, and cost and latency.

• Event-driven backend on AWS - FastAPI services in Docker, images on ECR, Lambda workers, EventBridge schedules and rules, plus queues and storage as needed, for jobs like weekly posting, review sync and audit runs.

• Webhooks and reliability - Receive and verify Meta and Google webhooks, handle retries, idempotency, rate limits, pagination and API version upgrades without breaking clinics.

• Observability and security - Logging, alerting and dashboards for every integration; least-privilege IAM; secrets in AWS Secrets Manager or SSM; careful handling of patient and clinic data.

• Ship with the team -Work closely with the Head of AI & ML and the growth team, write clear docs for each integration, and keep a short changelog of what broke and why.

Skills

• Language - Python 3, production-grade: typing, async, testing with pytest, Node.js /TypeScript for webhook or tooling work

• Framework - FastAPI, Pydantic, async HTTP clients (httpx/aiohttp), Celery or similar task queues

• Auth - OAuth 2.0 and OpenID Connect from memory: auth code + PKCE, scopes, refresh tokens, token security, Google service accounts, domain-wide delegation

• Google APIs - At least one shipped: GBP, Search Console or Google Ads, Google Analytics Data API, Places API

• Meta APIs - At least one shipped: WhatsApp Cloud API, Instagram Graph API or Marketing API
Embedded Signup, Conversions API, coexistence

• Platform approvals - Has taken an app through Meta App Review or Google OAuth verification to production, Business Verification, Tech Provider onboarding

• AI agents - Has shipped an LLM agent to real users: tool/function calling, structured outputs, multi-turn state, prompt design, Agent frameworks (LangGraph, LlamaIndex or similar), MCP, multi-agent workflows

• RAG - Embeddings, chunking, vector search (pgvector, OpenSearch, Pinecone or similar), hybrid search, grounding answers in sources, Rerankers, AWS Bedrock Knowledge Bases

• LLM ops - Evals, guardrails, tracing, cost and latency control across OpenAI, Anthropic or Bedrock models, Langfuse or LangSmith, fine-tuning, Indian-language handling

• Containers - Docker, multi-stage builds, docker-compose for local dev, Container image Lambdas

• AWS - Lambda, EventBridge, ECR, IAM, CloudWatch, SQS, ECS/Fargate, API Gateway, Secrets Manager, RDS/DynamoDB

• Data - PostgreSQL, schema design, migrations, Redis, DynamoDB

• Engineering habits - Git, code review, CI/CD (GitHub Actions), writing clear docs, Infrastructure as code (Terraform, CDK or SAM)

Role details

• Engagement - Full-time consultant, 3-month initial contract, renewable on performance

• Compensation - ₹1,00,000/month fixed

• Location - Remote · India

• Hours - Full-time, with overlap during Indian business hours; on-call for integration incidents shared with the team

• Equipment - Your own laptop; AWS, Google Cloud and Meta developer access provided

• IP and confidentiality - All code and work product belong to iris / Affinity Answers; NDA required