postgresql create index on join table/nested query to make it fast

postgresql(version 1.18.1) create index on join table/ or the way below is nested query

i have two table here:

CREATE TABLE users (
    id          SERIAL PRIMARY KEY,
    name        TEXT NOT NULL UNIQUE,
    role        TEXT NOT NULL,
    age     INTEGER NOT NULL,
    state   TEXT NOT NULL
);

CREATE TABLE sales (
    id          SERIAL PRIMARY KEY,
    uid         INTEGER REFERENCES users (id) ON DELETE CASCADE,
    pid         INTEGER REFERENCES products (id) ON DELETE CASCADE,
    quantity    INTEGER NOT NULL,
    price   INTEGER NOT NULL
);

right now i have the query to compute one state customer total sales

    select COALESCE(sum(quantity*price),0) as bigsum 
    from sales, users where users.role='customer' and users.id=sales.uid 
    and  state='New York'
"Aggregate  (cost=326562.33..326562.34 rows=1 width=8)"
"  ->  Hash Join  (cost=9927.33..325582.33 rows=196000 width=8)"
"        Hash Cond: (sales.uid = users.id)"
"        ->  Seq Scan on sales  (cost=0.00..163695.00 rows=10000000 width=12)"
"        ->  Hash  (cost=9682.33..9682.33 rows=19600 width=4)"
"              ->  Bitmap Heap Scan on users  (cost=424.32..9682.33 rows=19600 width=4)"
"                    Recheck Cond: (state = 'New York'::text)"
"                    Filter: (role = 'customer'::text)"
"                    ->  Bitmap Index Scan on index_user_state  (cost=0.00..419.43 rows=19600 width=0)"
"                          Index Cond: (state = 'New York'::text)"

where this give me 7940 ms. right now the sales table 10 million entries and users table 1000,000 entries i want it faster, how to do this? the index right now i have:

CREATE INDEX index_sales_user ON sales (uid);
CREATE INDEX index_user_id  ON users (id);
CREATE INDEX index_user_role_state ON users(role) WHERE role = 'customer';
CREATE INDEX index_user_state ON users (state);
CREATE INDEX index_users_role ON users (role) where role='customer';

i got the result Total query runtime: 7334 ms. 1 row retrieved.

 select COALESCE(sum(quantity*price),0) as bigsum 
 from sales  where uid in (select id from users where role='customer' and  state='New York')

"Aggregate  (cost=217188.73..217188.74 rows=1 width=8)"
"  ->  Hash Semi Join  (cost=9927.33..216033.22 rows=231101 width=8)"
"        Hash Cond: (sales.uid = users.id)"
"        ->  Seq Scan on sales  (cost=0.00..163695.00 rows=10000000 width=12)"
"        ->  Hash  (cost=9682.33..9682.33 rows=19600 width=4)"
"              ->  Bitmap Heap Scan on users  (cost=424.32..9682.33 rows=19600 width=4)"
"                    Recheck Cond: (state = 'New York'::text)"
"                    Filter: (role = 'customer'::text)"
"                    ->  Bitmap Index Scan on index_user_state  (cost=0.00..419.43 rows=19600 width=0)"
"                          Index Cond: (state = 'New York'::text)"