If the research of your ancestor stopped, say, at his approximate birth in 1690, what would be advisable to research in the notary deeds? If he was a person of some standing, he may have gone to a notary to give dowries to his daughters, so from when he was about 40 onwards we could begin to find some of these deeds. But also people with no great wealth used to go to a notary for the important passages in life.
Also, if you have an approximate date for his marriage and know the name of the wife, you could go looking for a dowry assignment from his father in law (the document could probably contain your ancestor's father's name). Finally, if you have a date for his death, he could have made a will (actually some men made more wills in the course of their life) giving many details on his family.
Archives have indexes of notary records indexed by notary names with the period they made deeds, and by place. Once you have found a notary for the period and place needed, you have to request the actual books and start to study them. A book may contain more years, but most typically at the end of each year there is usually an index, listing type of deed and one or both parties in the deed. In case the index is missing, the book is to be examined page after page, identifying the beginning of each deed and the names of the parties.
Most notary deeds were written in Latin, but Italian was used when the notary reported actual statements of the parties. With some practice, the researcher will easily identify the long legal formulas and jump to the content of the deed. Notary deeds will allow to identify people and relations among them, but will not give dates of birth or death, which would have to be approximated and derived from vital statistics.