Revenue runs on data. I build that data.
Marc Philip P. Turtal · Director, AI & Automations
I build the data and AI infrastructure revenue teams run on. 12+ years across fintech, e-commerce, and SaaS. One person delivering team-level scope by working AI-native.
JPMorganChase
My Amazon Guy
Payreto
Google
Telstra
DirecTV
Company | Confidential
A$208k
Revenue newly attributed to outbound DMs, 2026.
Confidential client
62%
Operational speed lift on an enterprise AI platform.
JPMorganChase
$51,568
Single-day funnel revenue from one webinar.
My Amazon Guy
800k+ EUR
Added monthly processing volume across two accounts.
Payreto
From zero to a revenue source of truth in 8 weeks.
Problem
Revenue data scattered across 8 tools, no attribution. Nobody could say which channel produced which sale, or how much cash a month actually collected.
Build
A Supabase Postgres warehouse on a three-layer architecture, fed by 12+ integrations: CRM, accounting, payments, paid ads, Instagram DM sales, Messenger, Discord, YouTube, call recordings, video analytics, webinars, and an intake form pipeline.
The warehouse now also carries paid spend and cash: a Meta Ads pipeline verified to the cent, and a payments feed tracking A$891k in running cash.
A read-only MCP server lets executives query the warehouse in plain English in Claude. Everything runs unattended on Railway crons.
Delivery
140+ PRs, delivered solo, AI-native. 14,297 Discord messages backfilled across 26 channels. 692 YouTube videos pipelined.
Stack in this build
Supabase
Postgres
Claude + MCP
Railway
Discord
YouTube
Instagram DMs
Messenger
21 sales
Fully tracked in the first complete month.
A$207k
Revenue visible end to end, by channel, matching the client's public figures 1:1.
A$82.5k
Cash collected, reconciled against payments.
The A$208k the founder could not see.
Problem
The client ran an outbound Instagram DM team for over a year with no hard ROI number. Nobody could say whether the DMs produced sales or just conversations. The team's budget was defended by anecdotes.
Build
Every Instagram DM conversation was pulled from the CRM into the warehouse and entity-resolved against buyers across the CRM, the sale form, and payment records. Each match was individually verified before it counted.
The result is a canonical attribution roster, rebuilt by a script rather than a spreadsheet, so the number stays alive as new sales land.
Delivery
45 DM conversations tied to closed sales, 36 of them started by the outbound team. A$208k in revenue attributed to specific conversations. The founder's first hard number on what outbound actually produces, now maintained automatically.
A$208k
Revenue tied to specific DM conversations.
36 of 45
Attributed sales started by outbound DMs.
100%
Matches individually verified before counting.
Then the warehouse started running the operation.
Problem
The data was unified, but operations still ran on memory. Commitments made in meetings and Slack died there. Nobody knew which sales calls were actually being recorded. Executives asked for numbers instead of receiving them.
Build
An AI operations layer on top of the warehouse. An AI triage desk reads meeting transcripts and Slack, extracts committed action items, posts them for one-click human approval, and files approved tasks into the task system. Tagging the bot in Slack files a task in about one second.
Call recordings flow in automatically, with coverage views that show exactly which reps record and which do not. A daily scorecard posts itself to Slack every afternoon, a weekly summary every Monday, and a live executive dashboard splits return on ad spend into cash collected versus contracted revenue. Error alerts and a daily health check watch the whole system.
Delivery
788 sales calls ingested and coverage-scored, which surfaced that nearly half of all sales calls were never being recorded, a people problem no one could prove before. Executive reporting now arrives unattended.
When a production incident hit, it was diagnosed and resolved the same day, and the core pipeline came out 30x faster (638 seconds to 21).
788
Sales calls ingested and coverage-scored automatically.
~1 second
From Slack mention to a filed, AI-extracted task.
30x
Core pipeline speedup shipped during a same-day incident fix.
Named companies. Shipped facts.
-
2026 to present
Director, AI & Automations
Company | ConfidentialA$208k revenue newly attributed to outbound DMs, A$891k in running cash tracked 1:1, an AI operations layer running triage, call coverage, and daily reporting unattended. 140+ PRs shipped solo. All three case studies above.
-
2025 to present
Senior Technology and AI Manager
JPMorganChase
62% operational speed lift from an internal AI platform. Project Ultron drove roughly 80% team efficiency gain and 30 to 40% faster incident resolution. Subpoena platform cut prep time 50%.
-
2023 to 2025
GTM & Marketing Automation Specialist
My Amazon Guy
70%+ conversion lift and 60%+ revenue growth from leading the order system migration to WooCommerce. $51,568 single-day webinar funnel. Company-wide AI adoption standard. 30% operational cost cut. Fastest Tier 1 to Tier 3 promotion in company history.
-
2020 to 2023
Technical Account Manager, Payments RevOps
Payreto
800k+ EUR monthly processing volume added by turning two underperforming accounts into the #2 and #3 revenue drivers. Led a team of 8. Managed enterprise clients including Delivery Hero (FoodPanda).
-
2012 to 2020
Earlier work
Google
Telstra
DirecTV
Consistently top KPI and CSAT with rapid promotions, plus dental practice ops and full-stack web delivery.
Built for founders who need revenue infrastructure, not headcount.
-
Data warehouse builds
A Postgres warehouse that unifies your revenue tools into one queryable source of truth.
-
Revenue attribution
Every sale tied to its channel, from first touch to cash collected.
-
AI access layers
MCP servers and Claude integrations so your team queries data in plain English.
-
Automation and bots
Unattended pipelines, AI triage desks, and bots that run your ops while you sleep.
-
Fractional AI and RevOps leadership
Director-level ownership of your data and AI roadmap, without the full-time seat.
-
New · Fixed-price websites
Website design and build
Fixed-price sites for small businesses, from a one-pager to a store. Eight design directions to pick from, three tiers, and you own everything at handover.
12 months free maintenanceYou own everythingNo lock-inSee packages and pricing at web.marcturtal.com
Things you can click.
Notes from the build.
- The triage desk that reads the meetings An AI layer that turns meeting transcripts and Slack chatter into approved, filed tasks, and why the human approval click is the whole design. Read
- Replacing a form with a checkout, and the four departments in the way Conversions up more than 70%, revenue up more than 60%. The engineering was the easy part; getting operations, finance, marketing and IT to agree on what a product is was not. Read
- The simplest RAG that works, and what a constrained environment teaches you Retrieval augmented generation in four steps, plus the design lessons that only appear when the model has limited permissions and a small context window. Read
- The payment tester that let non-engineers test payments Every tool for testing a payment API assumes you write code. Removing that assumption was worth more than optimizing the developers, and the gateway vendor adopted it as reference documentation. Read
- Shipping a revenue warehouse solo in 8 weeks The three-layer architecture, the verification rule that kept it honest, and why a one-person team can now carry a scope that used to need five. Read
- Giving executives a database without giving them SQL How a read-only MCP server turned a Postgres warehouse into something a non-technical founder queries in plain English, and the permission model that makes it safe. Read
Frequently asked.
Are you open to remote or international roles?
Yes. Based in the Philippines and built for remote work with US, AU, and EU companies. Comfortable with async workflows, overlap hours, and regular video syncs, any timezone.
Full-time employment or fractional engagements?
Both. Full-time Director or Head-level roles where I own the data and AI stack end to end, and fixed-scope fractional builds for founders, like the revenue source of truth case study on this page. Scope and terms are discussed per engagement.
What does a typical founder engagement look like?
A fixed scope with a working system at the end, not a slide deck. The 2026 build went from zero to a queryable revenue warehouse in 8 weeks, delivered solo. Most engagements start with attribution or a warehouse and grow from there.
Can I see your work or resume?
The case study and track record on this page are the portfolio. Public projects: phgiftideas.com and v2.mtits.me. For the full resume, and I approve each request personally, or reach out on LinkedIn.
Tell me what your revenue data should be doing.
- Emailcontact@marcturtal.com
- LinkedInlinkedin.com/in/mturtal
- Resume
- BasePhilippines · Remote, any timezone