What we build for research desks
One set of research agents that every fund needs, plus a pack for your strategy. Each deployment starts with one workflow and one analyst, runs in your cloud, and is live on your coverage in weeks.
What every desk gets
Agents that read what your fund already licenses, join it to your positions and coverage list, and report what changed since last time. Output goes where the team already works.
- Pre-open brief by email, per analyst or per book
- Broker notes compared with the prior note and consensus
- Earnings-call and filing comparisons, quarter on quarter
- Expert-call flags from your note-taker, with MNPI held back
- Notes filed to the right company in your notes system
- Model updates pushed into your Excel templates
What changes by strategy
The core stays the same. What changes is which documents matter and what "changed" means. A pack adds the sources and checks for your strategy.
| Strategy | What the pack adds |
|---|---|
| Fundamental equity long/short | Estimate revisions, call-to-call language shifts, alt-data reads against consensus, catalyst dates |
| Market neutral | The same checks across a wider universe, sorted by exposure so the brief stays short |
| Credit | Indentures and credit agreements, covenant headroom, debt footnotes, rating actions, maturity walls |
| Distressed and special situations | Court dockets, first-day declarations, claims registers, capital-structure changes, cleansing dates |
| Event-driven and merger arbitrage | Merger agreements, proxies, 13D/G filings, regulatory timelines across jurisdictions |
| Quant and systematic | Point-in-time text features delivered as code in your repository, ready for your backtests |
| Global macro | Central-bank statements compared meeting to meeting, data surprises, joined to your themes |
Data you already license
We work on top of your vendors, not instead of them. Pipelines check each release, compare vendors on the same KPI and store everything point-in-time, so nobody builds a signal on a rebased panel.
- Cross-vendor triangulation
- Panel health checks
- Cohort and share analyses, rerun on each release
- Vendor trials before you sign
- Entity resolution and ticker mapping
- Natural-language access to your warehouse
How engagements work
Fixed fee for the diagnostic and the pilot, agreed before we start. We'll give you a number on the first call.
Desk diagnostic
We sit with your team, map what they read and when, and build a working prototype of the most useful agent on your data.
Pilot
One or two agents live on real coverage in your cloud, tested against your own past examples, with code and documentation handed over.
Retainer
A forward-deployed engineer who keeps building with your team: new agents, new strategy packs, and upkeep of the ones you have.
Common questions
What is a forward-deployed engineer?
An engineer who works with the people using the software instead of from a distant backlog. They learn the workflow first-hand, build quickly and adjust to what they see at the desk.
Do our analysts have to learn a new tool?
No. Briefs arrive by email, notes land in the system you already use, and numbers go into your Excel templates. If you use Slack, Teams or Symphony, alerts can go there too.
Where does our data go?
It stays in your cloud account or on-prem. Model calls go only to providers you approve, under terms that forbid training, with zero data retention where the provider offers it. We can also use your own provider accounts, or models hosted entirely in your environment. See Security for the full terms.
Can you work with the data we already license?
Yes, and that's usually where we start. We build inside your environment and within your vendor agreements.
Do you share anything between clients?
No. General engineering experience carries over; your data, positions, prompts and outputs do not.
Do you work on-site?
Yes, when it helps. Most engagements are hybrid: in person for the first week and key milestones, remote with daily contact the rest of the time.
Which models and tools do you use?
Whatever fits your constraints. We work across the major frontier and open-weight models and build on your existing stack, such as Python, your data warehouse, Excel, Bloomberg and your OMS.
What happens when the engagement ends?
You keep the code, prompts, test sets and infrastructure, and we train your team to run them. Your data is returned and deleted, and we certify the deletion in writing. If you'd like us to stay on, the retainer covers that.
Will you sign an NDA?
Yes, before the first substantive conversation.
Start with a 30-minute call.
Tell us your strategy and what your team reads each morning. We'll tell you which pack fits and what a pilot would cover.