diff --git a/.github/workflows/codeql-analysis.yml b/.github/workflows/codeql-analysis.yml
index 85854f9..c81d3ba 100644
--- a/.github/workflows/codeql-analysis.yml
+++ b/.github/workflows/codeql-analysis.yml
@@ -24,6 +24,10 @@ jobs:
   analyze:
     name: Analyze
     runs-on: ubuntu-latest
+    permissions:
+      security-events: write
+      contents: read
+      actions: read
 
     strategy:
       fail-fast: false
diff --git a/cdk/lib/lambda/handler.py b/cdk/lib/lambda/handler.py
index fdac395..f99dbee 100644
--- a/cdk/lib/lambda/handler.py
+++ b/cdk/lib/lambda/handler.py
@@ -9,6 +9,8 @@
 from typing import Any
 from urllib.parse import parse_qs
 from botocore.exceptions import ClientError
+from datetime import datetime, timedelta
+from zoneinfo import ZoneInfo
 
 # Global cache for data
 _sessions_cache: list[dict] | None = None
@@ -35,11 +37,25 @@ def load_json_from_s3(key: str) -> dict:
 
 
 def get_sessions() -> list[dict]:
-    """Get sessions data with caching"""
+    """Get sessions data with caching and pre-parsed datetimes"""
     global _sessions_cache
     if _sessions_cache is None:
         data = load_json_from_s3(f"{DATA_PREFIX}/sessions.json")
-        _sessions_cache = data.get("sessions", [])
+        sessions = data.get("sessions", [])
+
+        # Pre-parse all session datetimes for better performance with SnapStart
+        # This happens once per Lambda instance and is cached across invocations
+        for session in sessions:
+            session["_start_dt"] = parse_session_datetime(
+                session.get("date", ""),
+                session.get("startTime", "")
+            )
+            session["_end_dt"] = parse_session_datetime(
+                session.get("date", ""),
+                session.get("endTime", "")
+            )
+
+        _sessions_cache = sessions
     return _sessions_cache
 
 
@@ -84,6 +100,75 @@ def parse_time(time_str: str) -> int:
     return 0
 
 
+def parse_session_datetime(date_str: str, time_str: str) -> datetime | None:
+    """Parse session date and time to datetime in Paris timezone"""
+    try:
+        # Parse date: "Nov 25, 2025" -> datetime
+        date_str = date_str.strip()
+        time_str = time_str.strip().upper()
+
+        # Parse time
+        parts = time_str.replace("AM", "").replace("PM", "").strip().split(":")
+        hours = int(parts[0])
+        minutes = int(parts[1]) if len(parts) > 1 else 0
+
+        if "PM" in time_str and hours != 12:
+            hours += 12
+        elif "AM" in time_str and hours == 12:
+            hours = 0
+
+        # Parse date and combine with time
+        dt = datetime.strptime(date_str, "%b %d, %Y")
+        dt = dt.replace(hour=hours, minute=minutes, second=0, microsecond=0, tzinfo=ZoneInfo("Europe/Paris"))
+
+        return dt
+    except Exception:
+        return None
+
+
+def get_paris_now() -> datetime:
+    """Get current time in Paris timezone"""
+    return datetime.now(ZoneInfo("Europe/Paris"))
+
+
+def filter_sessions_by_now(sessions: list[dict]) -> dict:
+    """Filter sessions happening now or starting soon (within 30 minutes)
+
+    Uses pre-parsed datetimes (_start_dt, _end_dt) cached in session objects.
+    This optimization leverages SnapStart caching for better performance.
+    """
+    now = get_paris_now()
+    in_30_min = now + timedelta(minutes=30)
+
+    ongoing = []
+    upcoming = []
+
+    for session in sessions:
+        # Use pre-parsed datetimes from cache (set in get_sessions())
+        start_dt = session.get("_start_dt")
+        end_dt = session.get("_end_dt")
+
+        if not start_dt:
+            continue
+
+        # If no valid end time, assume session is still ongoing if it started recently
+        if not end_dt or end_dt <= start_dt:
+            # Fallback: assume session lasts 20 minutes (median gap from analysis)
+            end_dt = start_dt + timedelta(minutes=20)
+
+        # Check if session is ongoing
+        if start_dt <= now <= end_dt:
+            ongoing.append(session)
+        # Check if session starts within 30 minutes
+        elif start_dt > now and start_dt <= in_30_min:
+            upcoming.append(session)
+
+    return {
+        "ongoing": ongoing,
+        "upcoming": upcoming,
+    }
+
+
 def filter_sessions(sessions: list[dict], params: dict) -> list[dict]:
     """Filter sessions based on query parameters"""
     filtered = sessions.copy()
@@ -201,31 +286,69 @@ def handler(event: dict, context: Any) -> dict:
 
     elif path == "/sessions":
         sessions = get_sessions()
-        filtered = filter_sessions(sessions, params)
-
-        formatted_sessions = []
-        for session in filtered:
-            start = session.get("startTime", "")
-            end = session.get("endTime", "")
-            time_str = f"{start} - {end}".strip(" -") if start or end else ""
-
-            formatted_session = {
-                "id": session.get("id", ""),
-                "title": session.get("title", ""),
-                "date": session.get("date", ""),
-                "time": time_str,
-                "stage": session.get("stage", ""),
-                "speakers": session.get("speakers", []),
-                "ecosystems": session.get("ecosystems", []),
-            }
-            formatted_sessions.append(formatted_session)
 
-        return create_response(200, {
-            "total": len(sessions),
-            "count": len(formatted_sessions),
-            "filters": {k: v[0] for k, v in params.items() if v},
-            "sessions": formatted_sessions,
-        })
+        # Check if 'now' parameter is present
+        now_param = params.get("now", [None])[0]
+
+        if now_param and now_param.lower() in ["true", "1", "yes"]:
+            # Filter sessions happening now or starting soon
+            now_filtered = filter_sessions_by_now(sessions)
+
+            def format_session(session):
+                start = session.get("startTime", "")
+                end = session.get("endTime", "")
+                time_str = f"{start} - {end}".strip(" -") if start or end else ""
+                return {
+                    "id": session.get("id", ""),
+                    "title": session.get("title", ""),
+                    "date": session.get("date", ""),
+                    "time": time_str,
+                    "stage": session.get("stage", ""),
+                    "speakers": session.get("speakers", []),
+                    "ecosystems": session.get("ecosystems", []),
+                }
+
+            paris_now = get_paris_now()
+
+            return create_response(200, {
+                "currentTime": paris_now.strftime("%Y-%m-%d %H:%M:%S %Z"),
+                "ongoing": {
+                    "count": len(now_filtered["ongoing"]),
+                    "sessions": [format_session(s) for s in now_filtered["ongoing"]],
+                },
+                "upcoming": {
+                    "count": len(now_filtered["upcoming"]),
+                    "description": "Sessions starting within 30 minutes",
+                    "sessions": [format_session(s) for s in now_filtered["upcoming"]],
+                },
+            })
+        else:
+            # Regular filtering
+            filtered = filter_sessions(sessions, params)
+
+            formatted_sessions = []
+            for session in filtered:
+                start = session.get("startTime", "")
+                end = session.get("endTime", "")
+                time_str = f"{start} - {end}".strip(" -") if start or end else ""
+
+                formatted_session = {
+                    "id": session.get("id", ""),
+                    "title": session.get("title", ""),
+                    "date": session.get("date", ""),
+                    "time": time_str,
+                    "stage": session.get("stage", ""),
+                    "speakers": session.get("speakers", []),
+                    "ecosystems": session.get("ecosystems", []),
+                }
+                formatted_sessions.append(formatted_session)
+
+            return create_response(200, {
+                "total": len(sessions),
+                "count": len(formatted_sessions),
+                "filters": {k: v[0] for k, v in params.items() if v},
+                "sessions": formatted_sessions,
+            })
 
     elif path == "/speakers":
         speakers = get_speakers()
diff --git a/llms.txt b/llms.txt
index 958eaec..53d6618 100644
--- a/llms.txt
+++ b/llms.txt
@@ -27,6 +27,13 @@ Returns conference sessions with server-side filtering.
   - Values: "morning" (before 12:00) or "afternoon" (12:00+)
   - Example: https://adoptai.codecrafter.fr/sessions?time=morning
 
+- `now` (boolean): Get sessions happening now or starting soon
+  - Values: "true", "1", or "yes"
+  - Returns: Sessions currently ongoing and sessions starting within 30 minutes
+  - Uses Paris timezone (Europe/Paris)
+  - Example: https://adoptai.codecrafter.fr/sessions?now=true
+  - **Note:** When `now=true`, other filters (date, stage, time, search) are ignored
+
 - `search` (string): Full-text search in title, description, speaker names, companies
   - Example: https://adoptai.codecrafter.fr/sessions?search=banking
   - Example: https://adoptai.codecrafter.fr/sessions?search=Anthropic
@@ -35,7 +42,7 @@ Returns conference sessions with server-side filtering.
 https://adoptai.codecrafter.fr/sessions?date=2025-11-25&stage=CEO%20Stage&time=morning
 https://adoptai.codecrafter.fr/sessions?date=2025-11-25&search=finance
 
-**Response:**
+**Response (regular filters):**
 {
   "total": 240,
   "count": 5,
@@ -59,6 +66,38 @@ https://adoptai.codecrafter.fr/sessions?date=2025-11-25&search=finance
   ]
 }
 
+**Response (with now=true):**
+{
+  "currentTime": "2025-11-25 10:15:00 CET",
+  "ongoing": {
+    "count": 3,
+    "sessions": [
+      {
+        "id": "...",
+        "title": "Session currently happening",
+        "date": "2025-11-25",
+        "time": "10:00 AM - 10:30 AM",
+        "stage": "CEO Stage",
+        "speakers": [...]
+      }
+    ]
+  },
+  "upcoming": {
+    "count": 5,
+    "description": "Sessions starting within 30 minutes",
+    "sessions": [
+      {
+        "id": "...",
+        "title": "Session starting soon",
+        "date": "2025-11-25",
+        "time": "10:30 AM - 11:00 AM",
+        "stage": "Mainstage South",
+        "speakers": [...]
+      }
+    ]
+  }
+}
+
 ### GET /speakers
 
 Returns all speakers with optional filtering.
@@ -143,6 +182,13 @@ user or found via web search), use this fallback strategy:
 
 ### Common Query Patterns
 
+**"What's happening right now?" / "What sessions are starting soon?"**
+→ GET https://adoptai.codecrafter.fr/sessions?now=true
+Returns:
+- Sessions currently ongoing (started and not yet finished)
+- Sessions starting within the next 30 minutes
+Uses Paris timezone (Europe/Paris)
+
 **"Find sessions about [topic]"**
 → GET https://adoptai.codecrafter.fr/sessions?search=[topic]
 Example: "Find sessions about AI in banking"