Senior Full-Stack Developer for AI-Powered Construction Project Management SaaS
Budget: $15 – $25 CAD
Senior Full-Stack Developer for AI-Powered Construction Project Management SaaS
Description:
I’m looking for an experienced engineer to join our team and take ownership of building out CPM-AI, a cloud-native, micro-services SaaS that ingests construction bid packages (PDFs, drawings, specs), applies advanced AI (OCR, RAG over LLMs, vision models, risk-scoring), and delivers structured scopes, interactive Q&A, and polished PDF/DOCX reports. You’ll implement the end-to-end pipeline—from file ingestion through AI micro-services, interactive FastAPI endpoints, budget optimization, to final report generation—with full observability, security, and audit-trail proofs baked in.
What You’ll Do
Build & Enhance Micro-services
FastAPI (Python) ingestion, OCR, parsing, classification, RAG scope extraction, risk scoring
Serverless/report-generator Lambdas with Jinja2 → LaTeX → PDF / Pandoc → DOCX
Front-End & API Layer
React/TypeScript UI for uploads, matrix views, report viewer
GraphQL-Yoga API gateway with tRPC hooks
Cloud Infrastructure & DevOps
Define IaC in Terraform + Helm charts for AWS EKS, RDS (Postgres + pgvector), S3, SQS, SNS, Redis, IAM
CI/CD pipelines with Bazel, Docker, ArgoCD, GitHub Actions, Blue/Green deploys, automated rollbacks
AI/ML Integration
Integrate AWS Textract & Tesseract fallbacks, fine-tuned LayoutLMv3, ViT-Large/16 + LoRA
Implement RAG retrieval with OpenAI Embeddings & pgvector; route to GPT-4o, Claude Opus, Gemini, Llama-3
Ensemble XGBoost/LightGBM risk-scoring service
Interactive LLM-Driven Assistant
FastAPI SSE endpoint /query for “What’s missing?”, “Who’s best for HVAC?” with inline ## Proof blocks
Auto-RFI drafting when confidence < threshold (SQS queue)
Security & Observability
JWT via JWKS, AWS Secrets Manager, VPC private subnets, WAF, TLS1.3, AES-256 at rest
Prometheus/Grafana metrics, Loki logs, Sentry traces, Great Expectations data QA
Must-Haves
5+ years Python (FastAPI) & TypeScript (React, GraphQL) production experience
Deep AWS expertise: EKS, Lambda, S3, RDS, SQS/SNS, IAM, VPC, Secrets Manager
Infrastructure as Code: Terraform + Helm/Kubernetes
LLM/AI integration: OpenAI API, embeddings, RAG frameworks (LangChain or custom)
Experience with OCR pipelines (Textract/Tesseract), spaCy NER, PyTorch vision models
Database & vector store: Postgres (JSONB), pgvector
CI/CD with Bazel, Docker, GitHub Actions, ArgoCD
Strong security mindset: JWT auth, encryption, compliance readiness
Nice-to-Haves
Experience with serverless PDF/DOCX generation (LaTeX, Pandoc, Tectonic)
Background in operations: Prometheus/Grafana, Sentry, Airflow orchestration
Familiarity with advanced ML tooling: LoRA fine-tuning, Llama-3, Claude, Gemini
To Apply, please send:
A brief summary of similar AI/SaaS platforms you’ve built
Your top 3 past projects involving micro-services on AWS
Confirmation you can start immediately and commit ~20 hrs/week
Description:
I’m looking for an experienced engineer to join our team and take ownership of building out CPM-AI, a cloud-native, micro-services SaaS that ingests construction bid packages (PDFs, drawings, specs), applies advanced AI (OCR, RAG over LLMs, vision models, risk-scoring), and delivers structured scopes, interactive Q&A, and polished PDF/DOCX reports. You’ll implement the end-to-end pipeline—from file ingestion through AI micro-services, interactive FastAPI endpoints, budget optimization, to final report generation—with full observability, security, and audit-trail proofs baked in.
What You’ll Do
Build & Enhance Micro-services
FastAPI (Python) ingestion, OCR, parsing, classification, RAG scope extraction, risk scoring
Serverless/report-generator Lambdas with Jinja2 → LaTeX → PDF / Pandoc → DOCX
Front-End & API Layer
React/TypeScript UI for uploads, matrix views, report viewer
GraphQL-Yoga API gateway with tRPC hooks
Cloud Infrastructure & DevOps
Define IaC in Terraform + Helm charts for AWS EKS, RDS (Postgres + pgvector), S3, SQS, SNS, Redis, IAM
CI/CD pipelines with Bazel, Docker, ArgoCD, GitHub Actions, Blue/Green deploys, automated rollbacks
AI/ML Integration
Integrate AWS Textract & Tesseract fallbacks, fine-tuned LayoutLMv3, ViT-Large/16 + LoRA
Implement RAG retrieval with OpenAI Embeddings & pgvector; route to GPT-4o, Claude Opus, Gemini, Llama-3
Ensemble XGBoost/LightGBM risk-scoring service
Interactive LLM-Driven Assistant
FastAPI SSE endpoint /query for “What’s missing?”, “Who’s best for HVAC?” with inline ## Proof blocks
Auto-RFI drafting when confidence < threshold (SQS queue)
Security & Observability
JWT via JWKS, AWS Secrets Manager, VPC private subnets, WAF, TLS1.3, AES-256 at rest
Prometheus/Grafana metrics, Loki logs, Sentry traces, Great Expectations data QA
Must-Haves
5+ years Python (FastAPI) & TypeScript (React, GraphQL) production experience
Deep AWS expertise: EKS, Lambda, S3, RDS, SQS/SNS, IAM, VPC, Secrets Manager
Infrastructure as Code: Terraform + Helm/Kubernetes
LLM/AI integration: OpenAI API, embeddings, RAG frameworks (LangChain or custom)
Experience with OCR pipelines (Textract/Tesseract), spaCy NER, PyTorch vision models
Database & vector store: Postgres (JSONB), pgvector
CI/CD with Bazel, Docker, GitHub Actions, ArgoCD
Strong security mindset: JWT auth, encryption, compliance readiness
Nice-to-Haves
Experience with serverless PDF/DOCX generation (LaTeX, Pandoc, Tectonic)
Background in operations: Prometheus/Grafana, Sentry, Airflow orchestration
Familiarity with advanced ML tooling: LoRA fine-tuning, Llama-3, Claude, Gemini
To Apply, please send:
A brief summary of similar AI/SaaS platforms you’ve built
Your top 3 past projects involving micro-services on AWS
Confirmation you can start immediately and commit ~20 hrs/week
Related categories:
Business, Accounting, Human Resources & Legal
Python
Ruby on Rails
Amazon Web Services
PostgreSQL