Alpha — pilots open

Labor units, calibrated to your work.

Hoursight turns your last twelve months of shipped jobs into tuned Accubid and ConEst labor units — so your next bid reflects how your crews actually build, not a national average from 1998.

Request pilot access Read the API docs →
Data connectors
6
Sage 300 CRE, Foundation, Spectrum, Vista, QBDT, Intacct
Estimating imports
2
Accubid Anywhere · ConEst IntelliBid
Tests green
178
engine · service · collector

How it works

  1. 01

    Point the collector at your accounting system

    The Hoursight on-prem collector reads job cost history from Sage 300 CRE, Foundation, Spectrum, Vista, QuickBooks Desktop, or Sage Intacct. Read-only. Nothing leaves your network unencrypted.

  2. 02

    We calibrate against your last twelve months

    The engine aligns each shipped job to its estimate, isolates the labor delta by assembly, and produces tuned units with confidence bands. Fuzzy match handles messy account structures; LLM matcher is optional.

  3. 03

    You get a signed report + Accubid/ConEst import

    Executive summary, per-assembly tuned units, a diff PDF against your last calibration, and a CSV that drops straight into Accubid Anywhere or ConEst IntelliBid. Import in fifteen minutes.

  4. 04

    Ongoing — monthly diff, webhooks, dashboard

    Every month we re-calibrate. You get a diff PDF, HMAC-signed webhooks into your project stack, and an ops dashboard that flags data hygiene issues before they poison a bid.

The API

Everything Hoursight does is exposed as an HTTP API. Bearer-token authenticated, HMAC-signed reports, request-scoped rate limits, an audit log for SOC 2.

Live status api.hoursight.ai (alpha) 
Loading…

Public

GET/healthLiveness
GET/readinessDB + worker + signing key
GET/metricsPlatform aggregates
GET/docsOpenAPI docs

Tenant-scoped (bearer)

POST/v1/uploads/{kind}labor_db · bids · job_cost
POST/v1/jobsSubmit calibration
GET/v1/jobs/{id}Poll status
GET/v1/jobs/{id}/reportSigned report PDF
GET/v1/jobs/{id}/diff.pdfDiff vs last run
GET/v1/tenants/{id}/dashboardHTML rollup
POST/v1/webhooksRegister webhook
GET/v1/audit-eventsSOC 2 audit log

Under the hood

Engine

Python calibration core. 83 tests. Fuzzy matcher + optional LLM matcher (Anthropic or OpenAI). Deterministic outputs for audit.

Service

FastAPI. 80 tests. Bearer auth, per-tenant rate limits, idempotency keys, HMAC-signed webhooks, request-ID trace, audit log.

Collector

Six on-prem adapters. 15 tests. Runs behind your firewall — nothing leaves without encryption.

Reports

ReportLab-generated signed PDFs. Executive summary, per-assembly deltas, hygiene warnings, month-over-month diff.

Pilot program

First five contractors: $4,500 one-time calibration + $1,250 / month ongoing. Regular pricing: $9,000 + $2,500 / month. Cancel anytime.

Request pilot access