Candid Intelligence Tech Stack

AI-powered pre-construction engineering automation for industrial projects

Construction San Francisco 2–10 employees Privately Held

Candid Intelligence applies large language models (Gemma, Qwen, DeepSeek) and agentic AI systems to compress months of pre-construction engineering into hours. The tech stack reveals a focus on document understanding (LangGraph, PydanticAI for structured extraction) paired with industrial engineering simulation (Aspen HYSYS, Aspen Plus), and the project backlog shows active work on safety validation, piping workflows, and scope-gap detection—suggesting the company is building toward fully automated design review and consistency checking for capital projects.

Tech Stack 23 technologies

Core StackPython PyTorch Neo4j React TypeScript Tailwind CSS Next.js FastAPI PostgreSQL Supabase Redis AWS Docker GitHub Actions Gemma Qwen DeepSeek LangGraph PydanticAI Aspen HYSYS Aspen Plus CI/CD Pydantic

What Candid Intelligence Is Building

◆Challenges

Reducing manual effort in preconstruction engineering
Automating repetitive workflow steps
Making preconstruction 10x faster
Pre-construction engineering collapse
Fast, inspectable, safe production tools
Large-scale document ingestion
Manual processing of engineering diagrams
Inefficient pre-construction workflows
Reducing safety review rework
Improving consistency checks

▲Active Projects

Mapping piping rework causes
Repeatable tagging and consistency checks
Product requirements for piping workflows
Secure execution backend
Feedback loops for ai improvement
Instrumentation and control workflow validation checks
Automating document understanding and consistency checks
Repeatable checks for scope gaps and inconsistencies
Agentic systems for pre-construction tasks
Human-in-the-loop flows for expert feedback

Hiring Activity

Accelerating20 roles · 10 in 30d

Department

Engineering

Product

Sales

Seniority

Senior

Intern

Company intelligence

Find more companies like Candid Intelligence by tech stack, pain points and active projects

Get started free

About Candid Intelligence

Candid Intelligence automates pre-construction engineering workflows for industrial construction and heavy-asset projects. The platform ingests engineering diagrams and specification documents, applies AI-driven analysis to detect rework causes and scope inconsistencies, and surfaces insights that typically require weeks of manual review by senior engineers. The team is senior-heavy (20 of 21 open roles at senior level) and engineering-focused (17 engineering, 2 product, 2 sales), based in San Francisco with active U.S. hiring. Current work spans document understanding, agentic task automation, human-in-the-loop expert feedback, and instrumentation validation.

HeadquartersSan Francisco

Company Size2–10 employees

Hiring MarketsUnited States

Frequently Asked Questions

What is Candid Intelligence's tech stack?

Python, PyTorch, Gemma, Qwen, DeepSeek, LangGraph, PydanticAI, Neo4j, Aspen HYSYS/Plus (engineering sims), Next.js/React, FastAPI, PostgreSQL, Supabase, Redis, AWS, Docker.

What is Candid Intelligence working on?

Agentic AI systems for pre-construction tasks, automating document understanding and piping workflows, safety review automation, scope-gap detection, and human-in-the-loop feedback loops for expert validation.