Back to the portfolio
Programme delivery
M&E System v2.0, FutureSense
A monitoring and evaluation system for the Peru and Panama hubs with three components: English proficiency in 5 levels, executive-function skills and perception. It includes redesigned field roles and print-ready instruments built for offline use and staff rotation.
Concept ✓
Next: Pilot in the 2026 cycle
- M&E frameworks
- Field instruments
- Offline-first
The problem
Programmes in Peru and Panama run on volunteer facilitators who can change every cycle, with little or no connectivity, and they still have to show donors a change that can be verified. Version 1 depended on a principal evaluator being present at every session.
What I built
I designed it.
Stages reached
Concept ✓
Result. Version 2.0, ready to pilot from the 2026 cycle.
Next: Pilot in the 2026 cycle
What the system measures
Three questions, asked in every programme cycle of 6 to 9 workshops.
| Component | Central question | Used by |
|---|---|---|
| English proficiency | Is the child moving forward in English? | Facilitators, programme delivery lead, hub, foundation |
| Executive function skills | Is the child building skills for life? | Facilitators, programme delivery lead, foundation |
| Perception | How do children and teachers experience the programme? | Programme delivery lead, hub, foundation, donors |
- Assessment is never presented as assessment: it happens inside normal workshop activities.
- If recording takes a facilitator more than 10 minutes after a workshop, the system is too complex. Every instrument is built for under 5 minutes.
- Facilitators do not interpret, they mark what they see. Interpretation happens at the end of the cycle.
- A simple fact recorded in 95% of sessions beats a rich one recorded in 50%.
- It works on paper in any condition. Nothing needs a connection until the end-of-cycle consolidation.
- It does not depend on any one person: a new volunteer can join in 30 minutes with the induction pack.
The trade-offs, written down
| What the system gives up | What it gets |
|---|---|
| Psychometric detail | Facilitators who actually fill it in |
| Perfect causal attribution | Evidence of change that can be defended |
| Sophisticated statistics | Data that is consistent between Peru and Panama |
| Deep individual diagnosis | Continuous group observation that others can repeat |
The cycle, from diagnosis to decisions
| Phase | What happens | Output |
|---|---|---|
| Before the cycle | An induction workshop for new people, a calibration of criteria with short case vignettes, and student IDs. | People ready and IDs assigned |
| 1. Diagnosis (sessions 1 and 2) | Four welcome activities that double as assessment, a starting English level from 1 to 5, a first reading of four executive-function skills, and a baseline of perception. | A starting profile for each student |
| 2. Follow-up (session 3 onwards) | A session card in under 5 minutes, an exit ticket at the end of every workshop and a weekly sync from paper to digital. | A running record by student and session |
| 3. Consolidation (end of cycle) | Two people review the cycle in 30 minutes, confirm or adjust levels, open or close alerts and pick one or two stories. | An updated profile, a list of alerts and story candidates |
| 4. Analysis (quarterly) | Three KPIs: level transition rate, alerts and exit tickets. A one-page note on what to adjust in teaching. | A quarterly dashboard and an adjustment note |
| 5. Use of results (90 days) | A two-page cohort snapshot for the foundation and donors, adjustments applied to the next cycle and decisions on expansion. | An external report and a plan for the next cycle |
Roles, redesigned
Version 1 depended on a principal evaluator being present at every session. The redesign moves routine work to a new role and keeps judgement with the evaluator.
| Role | What it does | Limits |
|---|---|---|
| Field facilitator (volunteer, may be new each cycle) | Runs the workshop, counts the exit ticket and fills the session card in under 5 minutes. | Only records. Does not confirm levels or raise alerts. |
| Certified recording facilitator (new role, one per school) | Checks the cards, moves data from paper to a sheet, assigns provisional levels and raises provisional alerts. Certified with a 60-minute workshop, two shadowed sessions and three calibration vignettes. | Cannot confirm a change of level, close an alert or publish an external report. |
| Principal evaluator (reduced role) | Trains and certifies, visits sessions 1 and 2 and at least one in three, and leads the end-of-cycle consolidation. | Holds the exclusive authority to confirm levels, close alerts and verify stories for external reports. |
| Programme delivery lead | Reviews the weekly sync, calculates the three KPIs each quarter and writes the adjustment note. | Acts as data guardian: checks records are complete and consistent before any analysis. |
An instrument: the exit ticket with three bins
One of nine print-ready instruments. It measures how a group felt when it left, and nothing else.
- Each child drops one token in a green, yellow or red bin on the way out. Three cups or paper bags on the wall are enough.
- It does not measure learning and it is not a grade. The facilitator counts the tokens in 30 seconds and writes three numbers on the session card.
- Alert rule: if the red bin is above 30% for four weeks in a row, the facilitator tells the programme delivery lead and does not change the workshop alone.
- Worked example from the instrument: 22 children, 15 green, 5 yellow and 2 red. 68% left feeling good and 9% feeling bad, so there is no alert.
Instrument 2 of 9, M&E System v2.0. The example comes from the instrument's instructions, not from programme data.
Product management and development skills shown on this page
| Skill | Where it shows |
|---|---|
| Stakeholder and role design | Six tiers, from the field volunteer to the foundation, each with what it decides and what it cannot. |
| Designing for staff rotation | A new volunteer starts in 30 minutes with a standard induction pack, and certification has fixed steps. |
| Choosing a few KPIs | Three: level transition rate, alerts and exit tickets. |
| Making trade-offs explicit | The plan states that it gives up psychometric detail to get real adherence from facilitators. |
| Process design with time budgets | A rule of under 10 minutes to record after a workshop, and an instrument for each moment of the cycle. |
| Instrument development | Nine print-ready instruments, from the session card to the calibration vignettes. |
| Reporting for donors | A two-page cohort snapshot every 90 days, with stories paired with the data of the same child. |