Early warning intelligence

The signals
were always
there

IT projects don't fail suddenly. They fail in slow motion — and the warning signs live in your team's communication data weeks before anyone raises a flag.

Join the waitlist See the signals
Communication frequency drop ◆
Slower response times ◆
Increased negative sentiment ◆
Communication silos forming ◆
Decision latency rising ◆
Key contributors going silent ◆
Cross-team threads collapsing ◆
Communication frequency drop ◆
Slower response times ◆
Increased negative sentiment ◆
Communication silos forming ◆
Decision latency rising ◆
Key contributors going silent ◆
Cross-team threads collapsing ◆
The problem
Your status reports are lagging indicators. By the time they turn red, the project has been failing for weeks.

We built project governance to track deliverables — scope, budget, milestones. But the conditions under which delivery becomes impossible go entirely unmonitored.

70%
of IT projects fail to meet their original goals
3–4
weeks earlier signals appear in communication data
Our approach

TenbitLabs is built on a research-backed premise: the human communication layer of a project is a more sensitive early warning system than any structured reporting process we have designed.

We analyse the patterns in how your team communicates — not the content of messages, but the structure, frequency, sentiment, and topology — to surface failure risk weeks before it becomes visible in formal reports.

This is not surveillance. It is signal detection.

Detection model

Four signals.
One early warning.

These patterns appear in communication data weeks before any formal escalation. They are quiet, consistent, and almost entirely invisible to traditional monitoring.

SIGNAL 01
Communication frequency drops
Overall message volume falls and contracts toward the core team. The project starts talking to itself. Edges go quiet first.
Leading indicator
SIGNAL 02
A B +40%
Response times stretch
Questions go unanswered. Decisions get avoided. People make assumptions instead of waiting. Each assumption compounds into risk.
Decision latency
SIGNAL 03
negative
Sentiment turns negative
Language shifts from collaborative to reactive. "What if we could" becomes "the problem is." The emotional reality of a failing project surfaces in words before it surfaces in reports.
Weeks before escalation
SIGNAL 04
Silos form inside the team
Sub-groups develop their own version of project reality. When those realities finally collide — at a milestone, a demo, a deadline — the gap can be severe.
Highest failure correlation
The system

How TenbitLabs works

A lightweight analysis pipeline that turns your project communication data into an early warning score — with the signals explained, not just flagged.

01
Connect your communication source
Export from Slack, Microsoft Teams, or email. TenbitLabs processes metadata and message patterns — not the content of your conversations.
Slack · Teams · Email · Jira comments
02
Signal extraction runs
The system analyses frequency patterns, response time distributions, sentiment trajectories, and network topology across your project's communication graph.
Research-backed detection model
03
Risk score + signal report generated
A project health score with the specific signals driving it — explained in plain language, with the trend over time. Designed to be shared with a project sponsor in a meeting.
Explainable, not a black box
04
You confirm or reject each signal
Your feedback trains the model. Over time, TenbitLabs calibrates to your organisation's communication norms — reducing noise, sharpening detection.
Feedback loop · Continuous improvement
About the lab

Built on research, not intuition

Research foundation
TenbitLabs is the product of an ongoing thesis on early detection of IT project failure using communication and behavioural data. Every detection signal is grounded in empirical research, not guesswork.
Building in public
The research, the product decisions, the things that work and the things that don't — all of it is documented in the blog. Follow the journey as it happens.
What we are not
TenbitLabs is not an employee monitoring tool. It does not track individuals. It detects project-level patterns in collective communication structure. The unit of analysis is the project, not the person.

Most project intelligence tools are built on structured data — tickets, velocity, budget burn, milestone status. That data is useful. It is also, by definition, a record of decisions already made and work already done.

TenbitLabs operates one layer earlier. Before a blocker gets logged in Jira, it surfaces as a thread that stopped getting replies. Before a team's alignment breaks down, it shows up as increasing communication silos. Before a project is formally at risk, the language people use to talk about it has already shifted.

We built TenbitLabs to make those earlier signals visible — and to give delivery managers something more concrete than instinct when they need to have a difficult conversation with a project sponsor.

The lab sits at the intersection of engineering, design, and intelligent systems. The research is rigorous. The product is being built for practitioners, not academics. The blog documents both.

Research blog

From the lab

The research, the product decisions, and the things learned along the way — documented in public as they happen.

Research
Your project isn't failing because of the technology — it's failing because of what your team stopped saying
5 min read
Research
What behavioural data can (and cannot) tell us about delivery health
Coming soon
Building
How my thesis became a product: the decisions that shaped TenbitLabs
Coming soon
Early access

Be the first to know

TenbitLabs is in active development. Join the waitlist to be notified when early access opens — and to follow the research as it progresses.

No spam. Unsubscribe any time. Early access notifications only.