Prepared for Gordon Law Group, P.C. — documenting mapped workflows, identified automation opportunities, projected ROI, and a recommended phased implementation plan for the IntelliFirm AI suite.
Over the paid discovery engagement (April 27 – May 18, 2026), Solve conducted structured sessions with Joshua Jewkes (litigation), Melissa (front-office), and Corbin Gordon (municipal law). Four workflows were observed, documented, and analyzed as automation candidates.
| Session | Participants | Focus Areas | Key Output |
|---|---|---|---|
| Apr 27 — Kickoff | Josh Jewkes, Melissa, Solve | Billing workflow, MyCase usage, pain points | Yellow-pad problem · "10/10" severity confirmed |
| May 1 — Billing Deep Dive | Josh Jewkes, Melissa | Time entry flow, rate structure, invoice approval | Crawl/Walk/Run scope defined; API requirements locked |
| May 7 — Process Review | Melissa | MyCase workflow, rate tiers, draft → approve → send | Billing automation requirements finalized |
| May — Litigation Session | Josh Jewkes, Solve | Discovery process, Bates numbering, case tiers | Document drafting automation scoped; Perry PRD initiated |
| May — Security Review | Corbin Gordon, Solve | Data handling, attorney-client privilege, AI policy | Security architecture requirements; BAA need confirmed |
Three high-impact automation opportunities were identified and prioritized from discovery. Each is addressed by a dedicated AI agent, sequenced to deliver fastest time-to-value while managing privilege risk and implementation complexity.
Root Problem: Attorneys rely on handwritten yellow-pad notes and deferred manual MyCase entry, introducing billing leakage rated "10/10" severity by Josh Jewkes.
Root Problem: Corbin Gordon operates in courtrooms where inbox monitoring is impossible. Urgent emails surface hours late with no automated urgency signal.
Root Problem: GLG generates ~75 legal documents per month requiring ~1.5 hrs of manual drafting each. Perry automates first drafts via RAG over the firm's own template library, with mandatory attorney approval.
For each of the three IntelliFirm products, this section documents the recommended technical stack, the explicit MVP scope (in and out), key assumptions, and the factors that could affect timeline or scope. Billy's scope is confirmed and ready for build. Miley and Perry remain preliminary pending the open questions surfaced during discovery — each is flagged in its own section below for resolution before formal SOWs are issued.
Miley is the highest-stakes product in the IntelliFirm suite from a privilege and confidentiality standpoint, because it reads attorney inboxes. The architecture below is intentionally conservative: classification, not comprehension; SMS metadata only, not email content; no data retained beyond classification outputs and audit logs.
Perry produces first drafts for attorney review and editing — never finished legal work. The attorney is always in the loop. This positioning aligns with current state bar guidance on AI use in legal practice and is the primary reason Perry's privilege exposure is lower than Miley's: Perry operates only when invoked, not continuously, and never sends, files, or finalizes anything.
Billy is locked. The architecture, scope, and assumptions above represent the formal Phase 1 SOW, ready for signature on engagement.
Miley and Perry remain preliminary. Each carries a short list of open questions — resolvable in a single 60-minute working session post-roadmap-acceptance. Once resolved, Solve will produce formal SOWs with specific timelines and acceptance criteria for each.
The three products share one foundation. Nuxt 4 + Supabase + Vercel/Cloudflare, with the auth, RLS, hosting, and PWA scaffolding established in Billy reused by Miley and Perry. That reuse is by design — and it's part of why the partnership pricing for the second and third products stays where it is.
SOW #1 commits to demonstrating a projected return of at least 200% on GLG's $17,000 cash investment. The analysis quantifies recovered value across all three products using billing rates and time savings confirmed during discovery.
| Product | Annual Value | GLG Cash Outlay | Solve In-Kind | Net Year-1 Return | ROI Multiple |
|---|---|---|---|---|---|
| Billy — AI Billing Assistant | $99,540 | $17,175 | $17,175 | +$82,365 | 5.8× |
| Miley — Email Triage | $77,025 | $15,560 | $15,560 | +$61,465 | 4.9× |
| Perry — AI Paralegal | $128,250 | $22,490 | $22,490 | +$105,760 | 5.7× |
| Full Suite — Combined | $304,815/yr | $55,225 | $55,225 | +$249,590/yr | 5.5× blended |
| Service | Purpose | Billy | Miley | Perry | Combined Est./Mo. |
|---|---|---|---|---|---|
| OpenAI / Anthropic API | AI inference | $25–50 | $15–60 | $30–120 | $70–230 |
| Supabase | Database, auth | $0–25 | $0–25 | $0–25 | ~$25 |
| Twilio SMS | Miley alerts | — | $5–25 | — | $5–25 |
| Vercel / Cloudflare | App hosting | $0–20 | $0–20 | $0–20 | ~$20 |
| Estimated Total Monthly Operating Cost (Full Suite) | ~$120–$300/mo | ||||
The phased build plan sequences all three products to deliver value incrementally while managing privilege risk and GLG's partnership investment. No sprint begins until the prior product is accepted.
| Phase | Product | Start | Target Completion | Duration | GLG Payment Milestone |
|---|---|---|---|---|---|
| Phase 0 | Discovery (SOW #1) | Apr 27 | May 26, 2026 ✓ | 4 weeks | $5,000 (paid) |
| Phase 1 | Billy | May 26 | June 18, 2026 | 3.5 weeks | $3,588 now + $8,587 on acceptance |
| Phase 2 | Miley | June 18 | July 23, 2026 | 5 weeks | $7,780 now + $7,780 on acceptance |
| Phase 3 | Perry | July 23 | Sept 17, 2026 | 8 weeks | $11,245 now + $11,245 on acceptance |
| Total Program Duration | ~20 weeks | $55,225 remaining GLG cash investment | |||
This engagement is structured as a co-development partnership. GLG and Solve jointly own the resulting IP through a Special Purpose Entity, enabling GLG to generate licensing revenue from the platform beyond their own practice.
Per Section 6 of SOW #1, the Parties will form an LLC to own all intellectual property from the full build. GLG's $5,000 Discovery Fee is treated as an initial Class A capital contribution upon SPE formation.
GLG total cash investment · Solve matched in-kind · Target completion September 17, 2026