01 / Research Copilot
Cite every claim back to the source.
Ask a question across coverage materials, filings, expert calls, and internal notes. Get an answer in seconds, with every claim cited to the exact page, paragraph, or table.
Learn moreCapabilities
Across every desk. One company brain.
02 / SOP Runbooks
Codify how your team works.
Capture the steps, structure, and voice behind an IC memo, EQ note, or CIM. New analysts produce on day one. Senior analysts stop reinventing every deliverable.
Learn more03 / Automated Slides
Decks in your firm's template.
Pitch books, IC decks, and initiation reports generated from your research. Each slide lands in your firm's template with your charts, formatting, and tone.
Learn more04 / Agents
Run analyses across your universe.
Describe an analysis you run by hand across hundreds of companies, deals, or assets. yAI builds the agent and executes it, returning a structured table.
Learn moreCaptured. Codified. Compounded.
How it works
Embedded engineering.
01
Embed
Our engineers embed on site with your team, learn how the work gets done, and configure your first workflow in week one.
→ runbook.yaml02
Run
yAI executes workflows against your firm's data. Every output cites the source and traces back to the original data.
→ memo.pdf03
Stay
We stay embedded as the firm grows, evolving the deployment with every new team, workflow, and model.
→ audit.logIntegrations
+ More
Customer story
“I ran an analysis and it saved me 2 weeks of work in 10 minutes.”
MANAGING DIRECTOR / LIFESCI CAPITAL
LifeSci Capital's bankers turn thousands of pages of trial data and literature into IC memos and pitch materials with yAI, each analysis running in minutes and every claim cited to source.
Read the full case studyEnterprise ready
Built for regulated firms.
Bring your data
Connect FactSet, PitchBook, internal repositories, expert calls, and your firm's warehouse. yAI builds the corpus.
factset · pitchbook · capiq · drive · sharepointStay in your cloud
Self-hosted in your Azure or AWS tenant. Data stays inside your perimeter. Model endpoints run per-tenant, scoped to your firm's data.
azure · aws · per-tenant · auditedEmbedded delivery
Engineers embed with each team to configure the first workflow in week one. The deployment scales across the firm under your governance.
firm-wide · engineers on-siteThe orchestration layer.
yAI is the orchestration layer. It routes per task across interchangeable models. A vendor price change, model deprecation, or term shift is a configuration update. The workflows, runbooks, and audit trail stay intact.
Research
Notes on retrieval, agents, and reliability.
Jun 29, 2026Agentic AI · Specification Gaming
The Specification Gap: Why Task Completion Is Not Intent Satisfaction
Even when an agent executes the stated task, uses the available tools, follows a plan, and produces a plausible result, the outcome may still fail the user's actual intent. This note introduces the Specification Gap: the structural gap between what the user means and what the agent operationalizes as a task.
Jun 8, 2026Agentic AI · Tool Use
The Tool Trust Problem: Why External Calls Do Not Automatically Ground Agentic Systems
External calls introduce signals, but they do not automatically introduce grounding. A tool call grounds an agent only if the selected tool is appropriate, the arguments are correct, the tool interface is reliable, the data is fresh, the output is valid, and downstream steps preserve the signal.
Jun 1, 2026Agentic AI · Verification
The Verification Paradox: Why Agents Cannot Automatically Validate Themselves
A verifier that shares the same information boundary, model priors, and error surface as the system it evaluates may increase confidence without increasing reliability. The most dangerous verification failure is not the one that looks broken—it is the one that looks fine.
Common questions
Where is yAI deployed?
yAI is deployed per tenant inside the customer's Azure or AWS environment. Customer data, embeddings, model traffic, and audit logs remain inside the firm's security perimeter at all times.
Does yAI train on our data?
No. Each tenant runs against an isolated model endpoint. Customer data is never used for model training, and model traffic is scoped entirely to the tenant. There is no shared corpus and no cross-tenant data flow.
Is yAI compliant with SEC 17a-4 recordkeeping requirements?
yAI is designed to meet SEC 17a-4 books-and-records requirements. Every prompt, retrieval, model response, and generated artifact is captured in an immutable audit trail, retained per the firm's policy and exportable on demand for compliance review.
What models does yAI use?
yAI routes per task across Claude Opus, Sonnet, and Haiku via Amazon Bedrock. Embeddings use OpenAI text-embedding-3-large. All inference is scoped to the customer tenant; no model traffic crosses tenant boundaries.
How are AI outputs verified and audited?
Every claim in a yAI output is checked against the source passages that support it. Admins receive a timestamped, filterable, exportable log of every data access and action. Deletions require two-step admin sign-off.
What deliverables does yAI produce?
Documents (research notes, coverage and sector reports, investment memos), slides (CIMs, pitch books, initiations of coverage), and tables (catalyst trackers, diligence trackers, CRM monitors). Each deliverable is a workflow that produces the artifact directly, with citation and firm voice enforced.
How long does deployment take?
Engineers embed with the first team in week one to configure the workflow. Deployment runs inside the firm's tenant over weeks one through four. Additional teams come online under your governance, with our engineers embedded throughout.