Example projectWeb appAICustom LLM
On-premises contract review
Hypothetical. Not a client, not a result.
// The problem
Reviewing a counterparty's paper against the firm's positions takes a senior reviewer hours, and juniors miss clauses the firm always pushes back on.
// What we would build
A custom model fine-tuned on the firm's own marked-up agreements and playbook, running on a GPU server in the firm's office or data centre so client documents never leave, that flags each clause against the firm's position for the reviewer to decide.
// What is in it
- A model fine-tuned on your past markups and your written playbook
- Each clause marked as accepted, flagged or outside your positions, with the reason
- Suggested fallback language from your own precedents, for the reviewer to accept or reject
- Measured against agreements your reviewers have already marked, before every release
- Runs on a GPU server you own, with nothing sent to a model provider
// Stack
- PyTorch
- Python
- Microsoft Word add-in
- PostgreSQL
- Lenovo GPU server on premises
// estimate
- Build
- ≈ US$137,000 to US$208,000, delivered within 31 weeksCAD 195,000 to 296,300
- Hosting
- ≈ US$176 a monthCAD 250 a month
- Support
- ≈ US$2,500 a monthCAD 3,565 a month
- Hardware
- ≈ US$47,800 to US$197,000CAD 68,000 to 280,800
Prices in your currency are estimates from today's Bank of Canada rate. All invoicing is in CAD or USD.
- AI route
- Custom LLM
- Model running cost
- ≈ US$176 a monthCAD 250 a month
Timeline by milestone
- Discovery
- 2.9 to 3.4 weeks
- Data and evaluation plan
- 0.4 weeks
- Design approved
- 2.4 to 3.9 weeks
- Core features
- 1.9 to 2.9 weeks
- Full build
- 1.4 to 1.9 weeks
- Training
- 6.4 to 11.4 weeks
- Testing and fixes
- 0.4 weeks
- Evaluation and red-teaming
- 3.9 to 4.4 weeks
- Launch
- 0.9 to 1.4 weeks
- Production
- 0.9 to 1.4 weeks
Outside our hands, and added to the calendar
- Hardware delivery, after it is ordered
- 2 to 6 weeks
How it is paid
- Deposit 20%
- CAD 38,760 to 59,020
- Discovery 1.5%
- CAD 2,907.00 to 4,426.50
- Data and evaluation plan 7.9%
- CAD 15,310.20 to 23,312.90
- Design approved 1.5%
- CAD 2,907.00 to 4,426.50
- Core features 4.6%
- CAD 8,914.80 to 13,574.60
- Full build 3.1%
- CAD 6,007.80 to 9,148.10
- Training 23.9%
- CAD 46,318.20 to 70,528.90
- Testing and fixes 1.5%
- CAD 2,907.00 to 4,426.50
- Evaluation and red-teaming 15.9%
- CAD 30,814.20 to 46,920.90
- Launch 1.5%
- CAD 2,907.00 to 4,426.50
- Production 8.6%
- CAD 16,666.80 to 25,378.60
- Holdback, 30 days after launch (10%)
- CAD 19,380 to 29,510
- GPU time for training and testing, at cost
- CAD 1,200. Billed up front, at cost, outside the milestones
- GPU server for AI, for the model
- CAD 68,000 to 280,800. Billed up front, at cost, outside the milestones