diff --git a/Makefile b/Makefile
index 79524d4f..168c8ff3 100644
--- a/Makefile
+++ b/Makefile
@@ -44,9 +44,7 @@ out/bhaagoindia.txt:
 # TODO: /exhibits.json is also helpful
 # And there are kn translations available as well.
 out/scigalleryblr.json:
-	# We only fetch events in the future
-	curl --silent "https://carbon-50388-default-rtdb.firebaseio.com/en/1ZJfGJT-7ZTOZoevdZZmh2hwXd2935ffJoWee9XXyFZ4/programmes.json" | \
-		jq '.[] | select(.timestamp | sub("\\.[[:digit:]]+"; "") | strptime("%Y-%m-%dT%H:%M:%S%z") | mktime > now)' > out/scigalleryblr.json
+	python src/scigallery.py
 
 out/venn.json:
 	python src/venn.py
diff --git a/README.md b/README.md
index 75113dc1..ae62fef5 100644
--- a/README.md
+++ b/README.md
@@ -11,7 +11,7 @@ We either have URLs that can be easily scraped (https://schema.org/Event), or a
 | Paytm Insider                  | ✔️      | ✔️     | ✔️          |                                                                                                           |
 | Bengaluru Sustainability Forum | ✔️      | ✔️     | ❌           |                                                                                                           |
 | Bhaago India                   | ✔️      | ✔️     | ✔️          |                                                                                                           |
-| Carbon Science Gallery         | ✔️      | ✔️     | ❌           |                                                                                                           |
+| Carbon Science Gallery         | ✔️      | ✔️     | ✔️           |                                                                                                           |
 | Max Mueller Bhavan             | ✔️      | ✔️     | ✔️          |                                                                                                           |
 | https://events.venn.buzz/      | ✔️      | ✔️     | ❌           |                                                                                                           |
 | https://linktr.ee/atta_galatta | ✔️      | ✔️     | ❌           |                                                                                                           |
diff --git a/src/event-fetcher.py b/src/event-fetcher.py
index 9c203cc7..4a9f1f9e 100644
--- a/src/event-fetcher.py
+++ b/src/event-fetcher.py
@@ -7,7 +7,8 @@
 
 EVENT_JSON_FILES = [
     'out/bic.json',
-    'out/sofar.json'
+    'out/sofar.json',
+    'out/scigalleryblr.json',
 ]
 
 KNOWN_EVENT_TYPES = [
diff --git a/src/scigallery.py b/src/scigallery.py
new file mode 100644
index 00000000..23a06991
--- /dev/null
+++ b/src/scigallery.py
@@ -0,0 +1,117 @@
+import http.client
+import json
+import datetime
+import re
+import os
+
+event_type_mapper = {
+    "Film": "ScreeningEvent",
+    "Lecture": "EducationEvent",
+    "Tutorial": "EducationEvent",
+    "Masterclass": "EducationEvent",
+    "Performance": "VisualArtsEvent",
+    "Workshop": "EducationEvent"
+}
+
+def guess_event_type(kind):
+    # search for each of the keys from the mapper
+    # in kind, and pick the first
+    for key in event_type_mapper:
+        if key in kind:
+            return event_type_mapper[key]
+
+    return "Event"
+
+tz = datetime.timezone(datetime.timedelta(hours=5, minutes=30))  # Asia/Kolkata timezone
+
+def fetch_data(host, path):
+    conn = http.client.HTTPSConnection(host)
+    conn.request("GET", path)
+    data = conn.getresponse().read()
+    conn.close()
+    return data
+
+def parse_duration(duration_str):
+    # Regular expressions to match hours, minutes, and days
+    hour_pattern = re.compile(r"(\d+(?:\.\d+)?)\s*Hours?")
+    minute_pattern = re.compile(r"(\d+)\s*Minutes")
+    day_pattern = re.compile(r"(\d+)\s*Days")
+
+    # Find all matches
+    days_matches = day_pattern.findall(duration_str)
+    if days_matches:
+        # If there are days, ignore hours and minutes
+        days = sum(int(match) for match in days_matches)
+        return f"P{days}D"
+
+    # Process hours and minutes otherwise
+    hours = sum(float(match) for match in hour_pattern.findall(duration_str))
+    minutes = sum(int(match) for match in minute_pattern.findall(duration_str))
+
+    # Convert fractional hours to hours + minutes
+    additional_hours, fractional_hours = divmod(hours, 1)
+    minutes += fractional_hours * 60
+    hours = int(additional_hours)
+
+    # Convert minutes to hours if 60 or more
+    hours += int(minutes // 60)
+    minutes = int(minutes % 60)
+
+    # Construct ISO8601 duration string
+    iso_duration = "P"
+    if hours or minutes:
+        iso_duration += "T"
+    if hours:
+        iso_duration += f"{hours}H"
+    if minutes:
+        iso_duration += f"{minutes}M"
+    return iso_duration if iso_duration != "P" else "PT0M" # Handle case with no duration
+
+def parse_timestamp(timestamp):
+    try:
+        return datetime.datetime.fromisoformat(timestamp).astimezone(tz)
+    # Return a very old date so this event is ignored
+    except:
+        return datetime.datetime(1900, 1, 1, tzinfo=tz)
+
+def make_event(e, ts):
+  experts = e["experts"].replace("_", " ").title().split(",")
+  return {
+    "@context": "http://schema.org",
+    "@type": guess_event_type(e["kind"]),
+    "name": e["name"],
+    "location": e["venue"],
+    "startDate": ts.isoformat(),
+    "description": e["blurb"],
+    "duration": parse_duration(e["duration"]),
+    "url": "https://carbon.scigalleryblr.org/programmes?p=" + e["uid"],
+    "performer": [{"@type": "Person", "name": expert} for expert in experts],
+    "maximumAttendeeCapacity": e["capacity"],
+    "organizer": {
+        "@type": "Organization",
+        "name": "Science Gallery Bengaluru",
+        "url": "https://carbon.scigalleryblr.org"
+    }
+  }
+
+def filter_data(data):
+    current_time = datetime.datetime.now(tz)
+    events = []
+    for p in data:
+        ts = parse_timestamp(p["timestamp"])
+        if ts > current_time:
+            events.append(make_event(p, ts))
+
+    return events
+
+def save_data(data, filename):
+    with open(filename, 'w') as f:
+        json.dump(data, f, indent=2)
+
+def main():
+    events = json.loads(fetch_data("carbon-50388-default-rtdb.firebaseio.com", "/en/1ZJfGJT-7ZTOZoevdZZmh2hwXd2935ffJoWee9XXyFZ4/programmes.json"))
+    filtered_events = filter_data(events)
+    save_data(filtered_events, "out/scigalleryblr.json")
+
+if __name__ == "__main__":
+    main()