Legacy load: High · Track: AI Adoption
IP Prosecution & Portfolio
Prosecution runs on deadlines and prior art, both of which are structured data wearing a law firm's clothes. adnah's read: automate the tracking, keep the claim strategy.
Ceremony, safe to remove
- Docketing deadlines by hand from office action letters that arrive in a dozen different formats
- Manually comparing a new office action's cited references against the portfolio's existing prosecution history
- Drafting boilerplate response shells before any substantive claim analysis has happened
- Re-summarizing the same patent family for internal status reports every quarter
Irreducible, stays with the lawyer
- Deciding how to amend a claim to survive a rejection without giving up commercial coverage
- Judging whether a reference actually anticipates or renders obvious, versus merely resembling the claim on the surface
- Portfolio strategy: which applications to keep prosecuting, which to let lapse, and why
- Client counseling on the tradeoff between claim scope and prosecution cost or delay
Primitives most in play
The usual failure mode
The recurring mistake is treating prior art search as fully delegable once a tool can retrieve references quickly. Retrieval is not the hard part of prosecution; reading a reference against the specific claim language and the examiner's actual objection is. Firms that let the tool's ranked list substitute for that reading end up filing responses that address the reference in the abstract but miss the examiner's real point, which costs another round and another deadline.
What changes
- Before: deadlines are docketed manually from each office action and cross-checked by a second person. After: deadlines are extracted and verified automatically, and the paralegal's time goes to the handful of ambiguous ones.
- Before: an associate builds a prior art summary from scratch for every response. After: the portfolio's own prosecution history and a first-pass reference summary are assembled automatically, and the associate starts from the claim analysis.
- Before: portfolio-wide status reports take a week of pulling matter files. After: the report reflects live docket state and the attorney's commentary is the only new writing required.
- Before: claim amendment language is drafted fresh each time. After: prior successful amendment patterns for similar rejections are surfaced, and the attorney chooses and adapts rather than starting blank.
What a working transformation looks like here
A working transformation shows up as a docket that nobody worries about missing, because the tracking has become mechanically reliable, paired with response quality that goes up rather than down, because the time saved on tracking and summarizing gets reinvested in reading the actual rejection more carefully. The portfolio review meeting becomes a strategy conversation instead of a status recitation.
Next practice area
IP Litigation →