Every altitude of delivery,
one honest system.
HEARTBEAT is a full project tracker — boards, plans, timelines, a one-minute daily app for the people doing the work — with an engine on top that compares reality to the plan every day and warns you while problems are still small and cheap to fix. Every suggestion shows its working, like a maths answer with the steps. (For the engineers among you: yes, it's a genuine control loop — the same math that keeps an aircraft level.)
The autopilot that reads drift before your Gantt admits it.
Every interval, the engine measures where reality sits against the plan, remembers how long the gap has been building, and tempers each correction with human factors. It produces one verdict and a ranked stack of actions — never a black box.
- PThis week's gap — how far reality sits from plan, right now.
- IThe memory humans lack — how long the gap has been accumulating.
- HpHuman factors — fatigue, morale, rework and honesty temper every move.
- ✓Guardrails are law — the overload cap can refuse overtime the engine would otherwise want.
#1 escalate sign-off to sponsor · unlocks 11.5 WTU · overtime refused — fatigue guard
Plan it, work it, verify it — in one place.
The autopilot only works because it sits on a real tracker your teams live in every day. Familiar to anyone who has used Jira, Asana or monday — with governance and field data built into the core.
Board
Swimlanes, WIP limits and guard-explained drag — moves that break a rule tell you why, live over SSE.
Backlog
Rank with drag or keys, group by epic, and move to a sprint with leave-aware capacity in view.
Sprints
Capacity that already knows who's on leave, with goals and burndown for cadence teams.
Timeline
WBS, dependencies and a critical path that the engine watches for slip — not just a static Gantt.
Grid
A spreadsheet view with inline edits that still pass every workflow guard. Bulk changes, safely.
Calendar
Due-date and milestone views for the people who think in weeks, not tickets.
A 45-second update from a crew leader with one bar of signal.
The signal that feeds the engine comes from the field. So the field app is offline-first, vernacular, and fast enough that people actually use it.
- ⚡45-second daily update — quantity, blockers, morale, done.
- ⛶Offline-first — capture without signal, sync truthfully when it returns.
- 🗣हिंदी / मराठी — vernacular by design, with a fallback chain.
- ◉Photo + GPS evidence — sealed to the log, visible in the web audit.
Control that holds up in an audit.
Approvals, change control and access aren't features you bolt on — they're how the platform behaves by default.
Approval gates
Baseline locks and milestone gates that only open when their conditions are actually met.
Change control
Every change request shows its quantified impact — before → after on end date, WTU and drift — before anyone decides.
Audit trail
Append-only, with changed-keys diffs and an access-explain tool that answers "why can't X see Y?".
Autonomy ladder
A0 → A2 autonomy with a type-to-confirm kill switch. You decide how much the autopilot may act.
Tenant isolation
Row-level security throughout — one organisation's data can never leak into another's.
Client portal
A status-words-only view for owners and clients, with sign-off and concern flows — and hard denials on anything sensitive.
The PMO layer executives actually read.
Roll every project up into an attention-ranked portfolio, plan capacity across people and weeks, and govern the front door so work enters on purpose — not by accident.
- ◆Attention score — the portfolio sorts itself by what needs you first.
- ▥Resource utilization — people × weeks, with overload flagged and rebalances suggested.
- ⇥Work intake — forms → triage → score → approve, with a WTU capacity check at the gate.
See it steer a real corridor.
A 30-minute demo on your own project shape — infrastructure, telecom, township or fab. We'll show the verdict, the arithmetic, and the guardrails.