A 30-minute call clarifies the question. We respond within one business day with a recommended starting point.
Start a conversationBuilt on twenty years of IP domain depth, an agentic engine called Lexa, and pod-based delivery that ships finished work product.
Trusted by the world's largest tech firms, licensors, and consumer brands — on the matters where the call has to be right.
Expert judgment where judgment is decisive.
Agent execution everywhere else.
That is what AI-native means at Dolcera.
Domain expertise and agentic AI built into the same delivery model — not stitched together, not bolted on. What follows is how that works.
An AI-native firm. Domain expertise and agentic AI built into the same delivery model — not stitched together, not bolted on.
Pod-based. One domain expert and Lexa per matter, in continuous dialogue. At scale, orchestrated loops with triage, where expert judgment is concentrated on the calls that matter.
Finished outcomes. SEP and EoU charts ready for licensing and litigation. Invention evaluations with defensible recommendations. NIA work. Patent landscapes that drive R&D investment.
Dolcera is one of a small number of firms with both deep IP domain expertise and a real agentic AI engine.
The domain. None of the AI.
Both. The contested ground.
Neither.
The AI. None of the domain.
That requires both layers — domain depth and agentic AI. Most firms have one. Dolcera has both.
Lexa is an agentic AI engine, purpose-built for IP workflows. The Dolcera team works with Lexa on every matter — orchestrating each as a sequence of decisions, directed by twenty years of expert IP domain depth.
Every matter pairs Lexa with a domain expert in continuous dialogue — agree, elaborate, or redirect. Twenty years of IP methodology guides the work.
A re-ranker scores result relevance on every query, tuned for precision at the confidence boundaries — where triage is decisive.
The same engine that produces one finished SEP chart screens ten thousand patents and surfaces the subset that warrants a closer look.
Client data is never used to train, fine-tune, or improve any model. Tuning runs against public benchmarks and public patent data — not client matters.
Each layer enables the layer above
Finished work product. SEP charts · EoU charts · NIA · invention evaluations · landscapes.
One domain expert and Lexa per matter. Continuous dialogue. Triage at scale.
Reasons through workflows. Single-matter mode and bulk mode. Re-ranker QA on every query.
Methodologies, taxonomies, and analytical playbooks. Refined matter by matter.
An AI-native firm requires all four layers. Pure tooling firms have layer 2 only. Pure human firms have layer 1 only.
The foundation.
Twenty years of patent prosecution, search, litigation support, and IP analytics — built into deep domain teams and an internal library of methodologies, taxonomies, and analytical playbooks. Refined matter by matter, drawn from public patent literature and public case history.
This is what equips experts to direct AI well in serious legal and strategic contexts.
The execution.
Lexa operates at two scales — depth on a single matter, breadth across a portfolio. Same engine, different cadence.
| Single-matter mode | Bulk mode | |
|---|---|---|
| Used for | SEP charts, EoU charts, invention evaluations, NIA work. | Portfolio screening, acquisition diligence, patent landscapes. |
| How it runs | Focused agent in dialogue with a domain expert. | Orchestrated loop across thousands of patents against a defined evidence pool. |
| Output | One finished, defensible deliverable. | Triaged subset for expert attention, or a synthesized landscape. |
The unit of work.
A pod is the smallest unit of work in an AI-native service. Default composition: one domain expert and Lexa, working on one matter, in continuous dialogue. The expert directs. Lexa executes. Review happens during the work, not after it.
Expert accepts the reasoning. Work moves forward.
Expert asks for more depth or stronger justification.
Expert re-directs Lexa to a different angle, evidence, or framing.
Across pods, Lexa's shared workspaces let a senior and junior analyst work the same matter with senior reasoning visible in real time — knowledge transfer happens during the work, not as separate training.
The pod is the unit. Lexa is the connective tissue.
In bulk mode, Lexa sorts each patent into one of three buckets, so expert time is concentrated where judgment is decisive.
Patents that don't need expert time don't consume expert time. Patents that need it always get it.
What the buyer receives.
Dolcera ships finished work product. The buyer uses it directly.
No client data trains any model. Period. Not now, not ever.
Lexa uses models trained on public datasets, paired with our internal methodology layer.
Tuning runs against public benchmarks and public patent data — not client matters.
Strict confidentiality clauses on every project. Data from one engagement does not flow to another.
Each one ships as finished work product. Buyers use it directly — counsel files the recommendation, licensing teams open the chart, R&D leadership acts on the landscape.
Prior art landscape plus a file / don't-file / file-with-modifications recommendation from a Dolcera evaluator.
References and report. Buyer's counsel makes the judgment call.
Standards-essential charting against 3GPP, multimedia codecs, Bluetooth, and Wi-Fi. Anchors licensing negotiations.
Patent-to-product mapping for assertion and litigation.
NIA evaluates the assertor's chart for technical gaps, stretches, and over-reads against products in light of the file wrapper. Invalidation searches prove lack of novelty.
Triage-based screening across full portfolios. Offensive mining surfaces hidden gems for licensing or assertion. M&A diligence assesses third-party portfolios for acquisition decisions.
Market intelligence from the global patent corpus. Run guided (testing a hypothesis) or unguided (surfacing patterns and trends).
Most engagements start with a 30-minute call. We respond within one business day with a recommended starting point.
There is no laboratory benchmark. The honest test is what holds up against human-only vendors and AI-only tooling on the matters that go to the wire.
EoU charts delivered for the world's largest software firm at velocity, on comparable scope.
EoU charts produced for the world's largest networking firm in a royalty dispute against a major audio and video standards licensor — in 2–3 weeks, after multiple other vendors failed to deliver usable work.
EoUs rated highly usable in a complex enterprise software litigation between Europe's largest software firm and a leading data analytics firm — relied on directly by counsel, no second pass.
Trusted vendor for Non-Infringement Assessment at the world's largest search firm and AI leader. Repeat selection on the hardest matters.
File-or-don't-file recommendations and invention-tweaking guidance for leading telecom licensing players and the world's largest search firm and AI leader.
Standards-essential charting against 3GPP, multimedia codecs, Bluetooth, and Wi-Fi for leading telecom licensing players. Anchors licensing negotiations valued at half a million dollars per firm and higher.
Hidden-gem discovery in major portfolios at the world's largest search firm and AI leader and the world's largest software firm. Unlocked cross-litigation wins and licensing value.
Third-party portfolio assessment and infringement risk evaluation feeding M&A decisions at the world leader in sleep and respiratory medical devices.
Shaping R&D investment and innovation strategy at the world's largest beverages and snacks firm, a global leader in oral care and personal care, a global leader in glass and building materials, and the world's largest eyewear firm.
Domain experts — not generalist analysts — across the technical fields where IP work is decisive. The people who scope the engagement are the people who direct Lexa during the work.
A 30-minute call clarifies the question, the decision behind it, and the timeline. We respond within one business day with a recommended starting point.
We respond within one business day.
