Best Sci score

from collections import defaultdict

def max_sci_score(travel, point):

    # build reward map
    reward = {p[0]: int(p[1]) for p in point}

    # build graph
    graph = defaultdict(list)
    for src, cost, dest in travel:
        graph[src].append([dest, int(cost)])

    # memoization
    memo = {}

    # perform dfs
    def dfs(node):
        """max score: reward - cost"""
        # base case: end point - no outward going node
        if node not in graph:
            return reward.get(node, 0)
        # use memoization
        if node in memo:
            return memo[node]

        best = -float("inf")

        for nbr, cost in graph[node]:
            candidate = dfs(nbr) - cost
            best = max(candidate, best)

        memo[node] = reward.get(node, 0) + best
        return memo[node]

    return dfs("start")

travel = [["start","3","A"], ["A","4","B"], ["B","5","END1"]]
point  = [["A","5"], ["B","6"], ["END1","3"]]
print(max_sci_score(travel, point))