One Live Decision. A Rigorous Framework. A Memory That Compounds.
Lantern Labs puts one live capital-allocation decision through the complete Lantern process, and keeps the resulting decision record current as the business and the valuation change. Pricing is published, not negotiated. Lantern Labs is the decision-intelligence system that helps PM-led family offices, emerging hedge funds, and concentrated funds turn one live investment decision into a rigorous, falsifiable, and compounding judgment. Lantern never tells you what to buy, sell, or hold — that decision stays yours.
What is the price actually for?
Lantern is priced against the value and the risk of the decision itself. Specifically, the price buys:
- Improving the quality of a consequential investment decision.
- Identifying the strongest argument against the thesis.
- Understanding what the current price already assumes.
- Reducing repeated reasoning errors.
- Preserving decision knowledge through time and personnel changes.
In short: you're paying for clearer, more confident decisions — not for a cheaper version of an analyst team.
Three clear tiers. One rigorous standard.
Published pricing means no negotiation. Choose the tier that matches the decision you need to make right now.
First Light
Best for: Testing the process on a name you're weighing right now.
- The full 14-dimension framework on one company
- The Weak Link and the strongest Anti-Thesis
- What today's price already assumes
- Falsifiers with thresholds and dates
- Confirmed / Inferred / Unknown on every data point
Updates available at $97 each.
Standing Watch
Best for: A concentrated book where every position needs the same rigor.
- Everything in First Light, across up to 15 names
- One decision record per name, updated as things change
- Top-down sector context alongside bottom-up economics
- Past misses carried into the next report's priority checks
- Updates included, priority turnaround
The Founder's Table
Best for: Making the decision record a permanent firm asset.
- Everything in Standing Watch, for twelve months
- Review sessions with Tejas Sarvaiya directly
- Your investment philosophy encoded into the framework
- An accountability ledger scored against outcomes
- Portfolio-level decision review
Limited to [X] clients per quarter.
15-minute fit call. No obligation.
In every report, at every tier
Against the cost of hiring
The Annual Partnership is roughly one-sixth the fully loaded first-year cost of a junior equity research analyst — and it produces work from day one.
That's the fully loaded first-year cost of a junior equity research analyst1 — paid in full from day one. Onboarding research finds that new hires in most roles take about twelve months to reach full performance.2
Lantern Labs is operational on day one.
What's the trust boundary — who owns my data and philosophy?
You own your data, your private context, and your encoded investment philosophy, and you can port it out. Lantern owns its framework, its system, and its pre-existing IP. Only generalized, non-confidential learning — never anything specific to your firm or your decisions — improves Lantern's reusable Sector Base. This directly answers the deepest, least-articulated anxiety in this buyer set: the fear that running a thesis through any AI system leaks a variant perception back into a model that quietly benefits everyone else. Lantern's answer is a real trust boundary, not a reassurance — your frame and decision history compound inside a perimeter you control, and any Lantern-side learning is consented and visible.
How does a Generic AI tool, an analyst-plus-AI setup, and Lantern actually compare?
| Capability | Junior analyst | LanternLabs® |
|---|---|---|
| Research quality | Variable | Standardized quality floor |
| Sector depth | Low initially | Varies by Sector Base |
| Firm context | Must learn | Encoded systematically |
| Framework consistency | Analyst-dependent | Organization-wide standard |
| Anti-Thesis discipline | Prompt-dependent | Mandatory workflow |
| Decision memory | Personal notes/files | Structured institutional asset |
| Continuity after departure | Low | Memory remains with the firm |
| Learning from mistakes | Usually informal | Classified and retrieved |
| Cross-analyst comparability | Weak | High |
| Management access and sector relationships | Limited | Usually weaker |
| Proprietary information | Limited | Limited unless supplied |
| Independent challenge | Limited by internal incentives | External and structurally adversarial |
Analyst columns assume LLM assistance.
What's the process?
- 01Read the framework and what a report contains.
- 02Recognize the problem: a thesis you can't fully defend, a memory that resets, knowledge trapped in one head.
- 03Choose a tier — a single report, ongoing coverage, or the annual partnership.
- 04Submit the decision: the name or ticker, and your investment philosophy — time horizon, what "quality" means to you, risk tolerance.
- 05Lantern runs the complete process and applies 100% human review.
- 06You receive the decision brief, the Anti-Thesis, what the price already assumes, and the falsifiers with thresholds.
- 07You inspect, challenge, or reject the reasoning. The final judgment stays yours.
- 08The decision record is updated as the business, results, and valuation change.
- 09Additional companies, sectors, or portfolio decisions enter through the same process — every unit of work remains a specific investor-side decision.
What information do you need from me before you can start?
Only: the one live decision (name, ticker, or deal); your investment philosophy (time horizon, what 'quality' means to you, risk tolerance); and how you currently track a thesis. Lantern deliberately does not ask for your full portfolio, book size, or position sizing — sizing and tax decisions are your Decision Making job, not Lantern's Decision Intelligence job.
Questions
How is this different from a research engine I already use for free?
Research engines are excellent and improving for free every quarter at search, retrieval, filings, and drafting a competent memo. What they don't do is remember your firm's philosophy between sessions, argue an Anti-Thesis with the same discipline every time, or build a decision record that's specifically yours. That's what a Lantern report adds.
Can I trust AI-generated analysis with a real capital decision?
Every output goes through 100% human review before you see it. The framework itself is fixed and public — you can inspect the 14 dimensions, the Anti-Thesis requirement, the falsifier structure, and the source-citation discipline before you pay anything.
Will you use my thinking or my data to help a competitor?
No. You own and can port your data and philosophy. Only generalized, non-confidential learning improves Lantern's reusable Sector Base — nothing specific to your firm or your decisions ever transfers to another client.
You're early-stage and founder-led — what if you're not around, or the memory moat never fully materializes?
That's a fair question to ask a young firm directly. Today, the moat is process quality and execution — the framework and the founder's judgment — which is documented and inspectable, not locked in one person's head. The compounding-memory advantage is real but still early; it's stated honestly rather than oversold.
I already have a process, or an analyst — why do I need this?
Most firms' processes are good instincts that are static, unenforced, and don't compound — they capture format, not judgment. Most analysts are excellent at coverage but represent a single point of failure: the reasoning leaves when they do. Lantern's reports and decision records make your own judgment — not a replacement for it — repeatable, inspectable, and durable.
Do you tell me what to buy, sell, or hold?
No, never. Decision Intelligence, not Decision Making. Lantern sharpens the decision; you make it.
Sources
- 1.Compensation figures reflect US equity research analyst salary data published by Indeed, combined with entry-level equity research compensation benchmarks. “Fully loaded” includes base salary, bonus, benefits, and employer overhead. Figure is stated conservatively; all-in cost for these roles commonly falls in the $130,000–$180,000 range.
- 2.Ramp-time benchmarks follow Gallup onboarding research, which finds that new employees in most roles take approximately twelve months to reach full performance, and recommends treating onboarding as a year-long process rather than a ninety-day one.
Lantern Labs is the decision-intelligence system that helps PM-led family offices, emerging hedge funds, and concentrated funds turn one live investment decision into a rigorous, falsifiable, and compounding judgment.
Book a Discovery Call