Lead Backend / Integrations Engineer (Python, FastAPI, Postgres)
First backend engineer at a US tax firm (based in Israel) building a client platform that automates US tax return preparation via the CCH Axcess REST API — intake, document extraction, e-filing, e-signature and payments. Stack: Python 3.12, FastAPI, Postgres, AWS, Stripe, Anthropic API; B2B contractor engagement.
- Compensation
- $3.5K–$6K / mo
- Location
- Not listed
- Work setup
- Remote
- Seniority
- Lead
- Employment
- Full time
- Experience
- 2+ years
- Category
- Backend
- Posted
Skills
Must have
- 5+ years building Python backends in production
- 2+ years as the person responsible for a system's reliability
- Deep Postgres: schema design, migrations, transactions, indexing, point-in-time recovery
- Integrations with unreliable third-party APIs: retries, idempotency, partial failure, long-running jobs with polling
- Security habits: least privilege, secrets management, no sensitive data in logs, dependency hygiene
- Docker; one major cloud; infrastructure as code
- Clear written English; documents decisions and writes runbooks proactively
- At least four hours of daily overlap with Israel (UTC+2)
- Availability December 2026 through June 2027 at 30-40 hours per week
Nice to have
- Fintech, payments, accounting, insurance or health-data experience
- SOC 2 or ISO 27001 exposure
- LLM API experience (Anthropic/OpenAI SDKs, tool use, structured outputs) or MCP servers
- Experience with XML-heavy enterprise APIs
- Russian or Hebrew language skills
Full description
AP US Tax Compliance and Advisory is a US tax firm based in Israel serving Israeli investors with US real estate and US citizens living abroad. Owner-run, growing, and building its own technology.
THE PROJECT
We prepare US tax returns in Wolters Kluwer CCH Axcess and have automated most of the preparation through CCH's REST API: import, calculation, printing, review memos, e-file qualification and release all run from code today. What runs in scripts must become a product before the 2027 filing season: a client platform where clients (or their AI assistants) complete a structured intake, the system extracts data from documents, prepares the return through CCH, routes it to a human reviewer and signer, collects an e-signature and payment, releases the e-file and delivers the result. The design document, data model, state machine, API and security model are written. You will be the first engineer and you will own the backend.
RESPONSIBILITIES
- Backend architecture and implementation: Python 3.12, FastAPI, SQLAlchemy/Alembic, Postgres, a job queue, workers, webhooks.
- The CCH Axcess API module as a server-side job runner: one in-flight job per return, global concurrency limits, exponential backoff on locks, idempotency keys, prompt handling, structured results, parked-job alerts.
- Security engineering: PII vault with field-level encryption (KMS), tokenized payload assembly, append-only audit log, RBAC with step-up MFA for sensitive actions, PII-free logging with a redaction layer, secrets manager, backups and restore tests.
- Integrations: Stripe (payments, webhooks), an e-signature vendor with identity verification, Google Drive, transactional email, WhatsApp Business API, Anthropic API for document extraction with human confirmation.
- Infrastructure as code on AWS (ECS Fargate or App Runner, RDS, S3, KMS, Secrets Manager, CloudWatch) or an agreed managed platform; CI with tests and security scans; staging and production environments; runbooks.
- You work against a mock CCH server built from our specification; production runs are performed by the firm. You will never hold production credentials or see client data.
- Collaboration: a full-stack engineer on the portal and staff console; an AI assistant that writes specifications and reviews every pull request before the owner sees it; a weekly 30-minute call with the owner.
- After the build: a retained maintenance agreement including on-call during the filing season.
REQUIREMENTS
- 5+ years building Python backends in production, 2+ years as the person responsible for a system's reliability.
- Deep Postgres: schema design, migrations, transactions, indexing, point-in-time recovery.
- Integrations with unreliable third-party APIs: retries, idempotency, partial failure, long-running jobs with polling.