Drop your IGC, get back the deepest tracklog analysis on the market — 100+ scalar metrics, 60+ charts, real DEM-based AGL, FFVL/XContest/CIVL scoring, plateau detection on five skill axes, and an AI coach that quotes your data, not generic theory.
No credit card · Free to try · Available in French & English
Standards & data sources we build on — from EN-B free flyers to PWC competitors
No instrumentation, no add-ons. Just your tracklog and a browser.
Drag-and-drop a single file or up to 50 in one go. Or hand over your XContest credentials once and let XCCoach scrape your entire history (years of flights) in the background. Automatic deduplication so the same flight never enters twice. We support every IGC variant in the wild — Skytraxx, Naviter, Flymaster, XCTrack, SeeYou, FAI standard.
Our analysis engine runs in the background: trims the flight to its first/last in-air fix, samples real terrain elevation per fix, classifies thermal/glide/float/prospect phases, scores XC, computes ~100 KPIs and 60+ charts — typically under 4 seconds for a 4-hour flight.
Open the flight detail page, scroll a methodically organised report, dig into any chart, read the Coach AI's debrief on your flight, share a read-only link with your coach, or export everything as a single self-contained PDF.
The dashboard is the page you actually open every morning — a curated mix of what's behind you, what's coming, and what to work on, refreshed silently as new flights are analysed.
Last five flights at a glance, plus year-by-year totals (hours, XC km, best score). Year selector to browse history.
Composite score from hours / 90 d (30 pts), flights / 30 d (20), SIV freshness (20), longest-recent XC and three rolling air-time windows. Concrete warnings when an axis is low.
Current week of your 4-week micro-cycle, completion %, and the highest-priority active goal with progress bar.
5-axis radar across your recent flights, side-by-side with monthly air-time bars. Two charts that summarise your season in two seconds.
Personal records broken in the last 90 d (longest XC, biggest gain, best vario, first triangle…) — the moments worth celebrating.
3-day outlook for your top site, scored 0–100 against your own "best-day signature" — the weather profile your good flights have in common.
Every flight produces 100+ scalar KPIs and 60+ charts. They're grouped into the categories pilots actually think in.
Total airtime, XC-circuit duration, tracker-gap-aware time accounting, ground-track distance, straight-line, best 60-min & 180-min rolling windows.
Free distance / Out & Return / Flat triangle / FAI triangle. Strict 5% closure check. League-specific multipliers (FFVL · XContest · CIVL). Sub-kilometre circuit precision.
Lift count, mean / max climb, vario consistency σ, mean gain per thermal, mean exit %, mean drift, turn-radius, bank-angle variability, time spent centred vs prospecting.
Mean glide ratio, mean sink, glide-efficiency vs polar (per glide), inter-thermal L/D, straight efficiency, floating share, prospecting share, transitions detected.
Task speed, XC ground speed, implicit MacCready, time on speed-bar, accelerated-time %, theoretical vs realised pace.
Max / takeoff altitude, cumulative gain (denoised), mean DEM-based AGL, max alt vs cumulus base, evolution stripe, baro-vs-GNSS source selection per metric.
Wind estimated from circling drift (time-weighted), wind alignment to circuit axis, transverse effect, circuit axis output %, distance lost to wind, off-wind XC distance.
High-resolution numerical weather forecasts merged from surface and pressure-level models. CAPE, T−Td spread, cumulus base, cloud cover, four-level wind profile, hodograph by altitude.
Net change of kinetic + potential energy summed per phase type, in absolute kJ, computed against your real flying weight (PTV — 95 kg default, or your account value).
Time below 50 m AGL, low-save count & lowest save altitude, sustained-sink events, worst rolling-window vario, low-save asymmetry — the strongest statistical predictors of incidents.
Five behavioural families — motor, decision, escape, breakpoint, context — covering cognitive load, reaction delay, bank variability, early-exit-low-alt, L/D-low-vs-high, min-accel-low and more. Catches fatigue and tunnel vision.
MacCready verdict, sector-by-sector speed analysis, deviation from theoretical, glide-efficiency-vs-polar per leg — for the days you fly with a task sheet in your hand.
Composite 0-100 score on five thermal-centering sub-skills — entry latency, vario stability, drift correction, turn symmetry and circle consistency. Measures how you worked the climb, not the air mass. Drives the centering axis of the skill radar.
Kilometres of XC per 1000 m of cumulative descent — a truthful conversion-efficiency metric. KPI card plus a 10-flight comparison chart so you can see whether you're getting more out of each metre of altitude over time.
A 0-100 score of the day's flying potential (CAPE, surface wind, cumulus base, cloud cover) plus the fraction you converted into XC distance. Honestly separates a strong-day 80 km from a weak-day 80 km.
For each climb, signed minutes you gained or lost vs the MacCready-optimal exit. Translates "MeanThermalExitPct" from an abstract % into a real-world XC-time cost.
Of all the times you started turning, what % became real climbs? Distinguishes pilots who turn on every whisper from those who commit late but well.
Km made-good per wall-clock hour, this flight + your 12-month median curve. Reveals whether you fade in the afternoon or take off too late, objectively rather than by feel.
Per-glide split of your realised L/D into glider polar baseline, wind component, air-mass lift, and line-straightness penalty. Separates "I flew well" from "I got lucky".
When you sink ≥ 200 m below your cruise altitude, how many minutes to recover? Median per flight + 12-month distribution histogram. A truthful proxy for tactical recovery skill.
And every numeric card carries a hover info icon explaining the formula and the source data — no black boxes.
The XCCoach analyzer is rebuilt and audited methodically. Every metric ships with a derivation note in the in-app glossary so you can see exactly how it's computed.
The Coach AI is wired into a state-of-the-art large language model with a structured digest of your real flight: phase ribbons, thermal-by-thermal details, climb & glide distributions, wind estimates, weather context, and your own progression history. It produces a one-shot, written debrief the way a club mate would — specific to this flight on this day.
Climbing · Gliding · Tactics · Wind exploitation · Safety. Every flight scores you 0–100 on these five pedagogical axes. When an axis stalls, the plateau detector flags it — offers a one-click 4-week training block that targets that exact weakness, and turns your recurring weak axes into tracked objectives you close over time.
Compares your last 5 flights to the prior 5 on each axis. Δ < 5 pts = plateau · Δ < 0 = regressing · < 20 flights total = early signal flag. 12-month window.
One auto-generated micro-cycle on your weakest axis. Weekly themes, deliberate-practice checklist, post-block retrospective comparing baseline → end.
Your recurring weak axes become persistent objectives — each records your baseline, sets a target and follows your progress across the next flights until it's reached and latched with a date. A focused shortlist of up to three, refreshed automatically. Deterministic — no AI, no cost.
Pick 2–4 flights, get a side-by-side table across nine sections (pacing, phases, wind, safety…), with best/worst highlighting and ±% vs median.
Anonymous, opt-in. k-anonymity ≥ 5 pilots required to render a value, your own flight excluded from the medians, mid-rank tie handling.
Hours / 90 days, flights / 30 days, SIV freshness, longest-recent XC, airtime in three rolling windows. Composite readiness score with concrete warnings.
Gamified milestones tied to real skill jumps — "Wind whisperer", "Axis master", first flat triangle, sustained 90-day climb-rate improvement, and 195 more.
Every climb you've taken across your full history, aggregated as a Leaflet heatmap. Four URL-bookmarkable filters (site / month / hour-of-day UTC / altitude band) so you can ask "where do I climb best in July afternoons above 2 000 m at Annecy?". The most-visited cores are also clustered (~500 m) and ranked in a sidebar — click a row to fly the map. Doubles as a Past climbs overlay in the 3D replay, so your usual cores light up while you debrief a fresh flight.
Recurring low saves are the strongest statistical predictor of serious paragliding incidents. XCCoach catches them — before they catch you.
Every section where the track dropped below 300 m AGL above real terrain AND climbed back ≥ 100 m. Start time, lowest AGL reached, duration, altitude regained — with a "Show on map" button.
Total minutes spent in the critical danger zone — DEM-based, not a flight-floor proxy. Red as soon as > 0. Surfaces in the per-flight KPI strip and in the multi-flight comparison.
Worst rolling-window vario (baro-preferred), longest sustained-sink segment, low-save asymmetry (do you always turn the same way when low?), recovery climb tracking.
When an OpenAIP key is configured, the analyser checks the track against published airspace and flags violations with timestamp, altitude, and class.
20 indicators across 5 behavioural families — cognitive load, reaction delay, bank variability, early-exits-low-alt, L/D low-vs-high. Helps catch fatigue patterns before they cost you.
Repack reminders, last-SIV freshness scoring (capped at 10/20 if no date recorded), reserve inventory with date-of-purchase and next-repack dates.
A dedicated view that aggregates safety signals across all your flights — KPI cards for rolling 30 / 90 / 365-day windows (low saves, sub-50 m AGL exposure, airspace warnings), behavioural stress trend chart, top-3 risk flights with one-click drill-down, and a per-site airspace-violation table. The single screen that tells you whether things are getting better or worse.
A high-precision route optimiser finds your true best circuit (start point not pinned to the take-off), with sub-kilometre accuracy and a strict 5% closure check on triangles. Pick your league once — every flight rescores at the next analysis pass.
Best route across the flight — start point not pinned to the take-off.
Out + back leg sum (not 2× farthest). Multiplier per league.
No leg > 45% of the total. ×1.2 (FFVL).
Each leg ≥ 28% of the total. ×1.4 (FFVL). Strict closure.
From "is the day worth driving for?" to "what's my safe height on leg 3?" — every layer of pre-flight planning, on one screen.
Each upcoming day scored 0–100 against your best-day signature — derived from your own historical good flights, not a generic threshold table.
06:00 → 21:00 UTC: surface temp, wind, gusts, wind at four pressure levels, CAPE, cloud cover, rain — cells turn amber then red above pilot-safe thresholds.
Hourly wind + temperature profiles for the day, plus daily convection indicators (cumulus base, BLH proxies, lapse rates) sourced from a high-resolution numerical weather model.
Synthetic per-hour-per-location thermal score derived from surface + pressure-level outputs. Quick "when does the day fire?" overlay on the timeline.
Open the site detail page to get the same forecast bundle aimed at that exact spot: wind rose, hourly thresholds, soundings, flyability, plus your own history at the site.
10 m / 850 / 700 / 500 hPa wind on every forecast view, with shear highlighting between layers — the data you'd otherwise scrape from three different websites.
Three-period evolution (11h / 14h / 17h UTC) with shear highlighting. Head/tail component along each leg, crosswind magnitude.
When a glider is bound: required AGL at the wing's finesse + 20% margin. Bucketed easy / moderate / hard / stretch — a "stretch" leg needs more than 1500 m above ground.
Every named site within 60 km of your anchor: distance, compass bearing, straight-line AGL needed at the wing's finesse + 20% margin.
Terrain sampled every ~5% along every leg from a digital elevation model, drawn as one continuous chart. Combined with the glide requirement: minimum flight altitude per leg.
When OpenAIP is configured, the planner draws restricted airspace on the route map and surfaces the ones a leg would skim, before take-off.
Persist named XC plans and attach them to a flight after the fact for plan-vs-actual analysis.
Beyond the flat list, XCCoach surfaces your career as a graph of sites, gliders, milestones and seasons.
Every take-off you've used, auto-grouped by ~110 m grid cell so "St-Hilaire" is one site even when you tagged it three different ways. Add your own named sites with notes & orientations.
Per-site: history of flights, weather forecast, your best XC there, contextual percentiles, wind rose across all your visits.
Learn each site like a local: recurring "signature" thermals mined from your full history, the weather profile of your best days there, your personal envelope vs today's forecast, and percentile bars ranking you against the global cohort.
Chronological flight log with rich filters: by site, by glider, by year, by tag. Bookmarkable URLs, every column sortable.
Long-form journal view that combines per-flight notes, conditions, learnings into a single narrative — exportable as a printable PDF for your annual self-debrief.
"Year in review" page: hours, XC km, best score, top sites, biggest improvement axes, retroactive PRs and notable flights.
Bind your wings, record their trim airspeed. Get per-glider performance breakdown — which wing climbs better in your hands, which one glides further per Δalt.
Auto-derived per-flight badges in the list view: "New PR", "First triangle", "Longest of month", "Highest of year" — surfaced exactly where you scan for them.
Track gear maintenance: line-trim checks, reserve repacks, instrument calibration — with date-driven reminders before they're overdue.
Multiple reserves per pilot, each with serial, manufacture date, last-repack date and next-due alert. SIV freshness feeds back into the readiness score.
A separate workflow for competition pilots: import the task in XCTrack format, plan it against forecast wind, fly it, and get a sector-by-sector debrief that quantifies where you lost time on the round.
Placeholder — replace with screenshot of /competition
Beside the AI debrief, XCCoach runs a deterministic recommendations engine that produces a prioritised, reproducible improvement list — the same input always yields the same output.
A short ranked list of actionable improvements per flight (and across the last N flights), each with the underlying KPI that triggered it. Reproducible: same data → same advice.
For each active goal you've set, a dedicated diagnosis: how close are you, what's blocking you, and which next flight should target which axis to unblock.
The Coach AI debrief is generated next to the deterministic recommendations — same flight digest, two complementary readings: a reproducible list, and a written narrative.
Every metric ships with a derivation note in the in-app glossary.
Full i18n — 4500+ keys with strict EN/FR parity. The UI follows your browser language and remembers your override.
Generate a read-only token URL for any flight. Your coach reviews the analysis without an account, no PII leaked, expires when you say so.
Print-optimised export of every flight, with three presets: Short (1-2 pages KPIs + map), Standard (adds safety, notes, comp, weather) and Full report (every chart). Embedded charts, no external assets needed.
One-click XLSX export of your entire flight history in the FAI logbook format — ready for licence paperwork or your national federation.
Replay any flight on a real-elevation 3D globe (MapLibre GL + deck.gl) with satellite imagery, animated trail, the pilot's name floating with live ground speed & L/D, billboarded wind arrows revealing one-by-one, a translucent glide-reach cone, 3D extruded airspaces colour-coded by ICAO class, and a Past climbs overlay that drops dots where you usually find lift in the area. Time slider, 1× → 120× playback, chase-cam, three camera presets (top / wide / side at 85° pitch), and one-click WebM video export.
Installable on iOS, Android and desktop. Offline-ready shell + cached analyses so last week's flight stays reachable without signal.
Set a target (first 100 km, first FAI triangle, 90-day air-time goal). XCCoach tracks progress on the dashboard and feeds the goal-aware coaching with each new flight.
Verification, password reset, alert notifications, announcement emails — all routed through a configurable SMTP relay.
Each upload is fingerprinted against your previous IGCs — the same file can't enter twice, even under a different name.
When the analyzer evolves, every flight in your history can be re-analysed with one click — your numbers always reflect the current methodology.
Honest dark mode across every page (charts, tables, maps included). Respects your OS preference on first visit, persists your override.
"I stopped flying blind. The plateau page told me my circuit-axis exploitation hadn't moved in two months — three flights of focused practice and the radar shifted."
"The Coach AI caught a pattern across five flights that I'd missed: I was bailing thermals 200 m below my own median. One sentence, three weeks of skill gain."
"I asked my school to share flights this way. The PDF export is the cleanest debrief support I've seen — and the multi-flight comparison nails what students actually need to look at."
Start free. Upgrade when you want unlimited uploads and the Coach AI.
Try the full analysis pipeline.
For pilots who fly weekly.
Schools, federations, competition teams.
Free to try, no card needed. We'll show you what your tracklog has been hiding.