Manager of Applied AI Architecture, Startups · San Francisco, CA · Hybrid
Lead the startup architects —hire, coach, still ship.
Manager of Applied AI Architecture, Startups leads and coaches the pre-sales architect team, personally carries strategic early-stage accounts, and owns Claude playbooks for the Developer Platform. Fauzul's spine here is people leadership and founder GTM — EM → CIO, 22-person core, alumni leading abroad — plus 0→1 agentic builds (, , ) as the technical bar he would model. Claude CMS at is light builder credibility, not the deep band (Zürich owns that showcase). SF is a deliberate relocation into the seat.
- People leadership
- Startup GTM
- 0→1 credibility
- Playbooks & evals
Manager of Applied AI Architecture, Startups: hire and coach architects, carry strategic early-stage accounts, own eval playbooks, partner with New Business — people leadership first, builder credibility required.
How the work maps
What this manager seat actually needs
Not Beneficial Deployments' mission lane and not an IC Claude-CMS clone — a player-coach who grows SA talent and still opens a repo.
Grow SA talent from delivery leadership
EM → CIO growth of a 22-person core and alumni leading abroad — people systems ready to translate into SA hiring bars and deal coaching.
0→1 peer credibility
, , as the deep builder bar — full mechanism lives once in the proof band below.
Evals the team can repeat
Context engineering and evals against real workflows — model the playbook on strategic accounts, then scale it through coaching.
After joining
First solutions as Startups Architecture Manager
Translate EM/CIO systems into SA rituals fast — while staying personally on a few strategic early-stage accounts.
Write the hiring bar in the first two weeks
Define builder credibility screens: open a repo, frame an eval against a founder's workflow, say no when a pattern will not hold. Model with // receipts.
Own two strategic startup accounts personally
Player-coach from day one — deal coaching visible to New Business AEs while the team shadows, then inherits playbooks.
Publish the first eval playbook draft
Discovery → eval → competitive bake-off checklist for Claude on the Developer Platform — living doc the team can run without bottlenecking on the manager.
8+ years in technical customer-facing roles (Solutions Architect, Sales Engineer, Forward Deployed Engineer, or similar).
10 years co-founding as the customer-facing technical lead — discovery, architecture, POC, and go-live with startups and enterprises across nine geographies. Same work as SA/FDE under founder titles.
5+ years leading and managing pre-sales or technical go-to-market teams; track record hiring and coaching high-performing SA teams.
Engineering Manager (2022–2024) then CIO — grew to a 22-person core (52+ with contractors), hired across engineering/design/QA/product, coached alumni now leading in Norway (2), Canada and Germany. Led customer-facing delivery pods, not a formal vendor pre-sales SA org.
AdaptTranslate player-coach delivery leadership into SA-team rituals: hiring bar for technical credibility, deal coaching, eval playbooks, and paired work with New Business AEs. The people-growth muscle is proven; the titled GTM-SA manager track is the ramp.
Deep experience with startups or high-growth technology companies — velocity, constraints, and how technical decisions get made.
Lived the early-stage constraint set as co-founder: weekly ship cadence, resource limits, founder/engineer dual audience. Advises and builds with startup teams (, , ) and understands seed-to-growth procurement instincts from the seller and builder side.
Genuine builder credibility — shipped LLM-powered applications; earn trust of technical founders and founding engineers without relying on a title.
, (Milan AI Week), (Google × Kaggle Agents Intensive) as 0→1 agentic receipts under startup constraints. Daily Claude Code / Claude Design; Anthropic Claude Code 101 and Introduction to Agent Skills certifications.
Hands-on expertise with context engineering, LLM evaluation frameworks, and modern AI architectures; comfortable with Python and common LLM tooling.
Daily Python and TypeScript builder. Context budgeting, MCP resource selection, multi-agent routing (A2A), and evals framed against real workflows until systems hold in production — not only in demos.
Partner with New Business sales leadership; surface startup build patterns as Product and Engineering feedback; build presence via events and content.
Co-founder commercial partnership — priced deals, scoped SOWs, ran workshops and public teaching (@FrenchwithFauzul, developer sessions). Field patterns already become Skills and harness choices. Has not owned a formal vendor GTM segment strategy with AE leadership inside a frontier lab.
AdaptReuse founder–AE pairing instincts and public teaching muscle; learn Anthropic's startup segment operating cadence and AE rituals in the first weeks while staying personally on strategic accounts.
Passion for making powerful technology safe and societally beneficial.
Treats compliance as a product surface — HIPAA at (3,242 patients · 13,340 schedules · 2 yrs 8 mo), GDPR/VAT at (acquired by Loopcloud 2026). Chose Claude for production predictability rooted in Anthropic's safety-first DNA.
Top strength
People growth with receipts
EM → CIO; 22-person core (52+ with contractors); alumni leading in Norway, Canada and Germany — coaching and hiring proven for an SA team.
Also day-one
Supporting strengths
0→1 bar founders inspect
, , and — 0→1 agentic receipts under real startup constraints; full builder walkthrough is in the proof band below.
Technical founder GTM signal
The posting calls out founder-led sales motion as a strong-candidate marker — priced deals, scoped SOWs, sat with technical buyers while still shipping.
Adaptable gaps
Named openly, with a bridge
Not 5+ years as a titled pre-sales SA manager inside a vendor GTM org.
Map EM/CIO player-coach systems onto SA hiring bars, deal coaching, and eval playbooks. Stay personally on strategic startup accounts while the team scales — the JD already expects that hands-on model.
San Francisco is not his current base.
Hybrid ~25%+ office commitment from day one on Pacific hours. Sister in London and European delivery history keep him a relocator, not a remote-from-Dhaka arrangement.
Production proof
The builder bar a startup architect team can copy
I do not want an IC escape hatch — I want architects who can open a repo, sit with a founder, and say no to a Claude pattern that will not hold. That bar and the people system to hire for it are already proven; the titled SA-manager track is the named ramp, not first-time leadership.
Fauzul · Manager, Startups Applied AI Architecture
TagRamp · NewScriber · VisaPros — 0→1 under startup constraints
The deep proof for this manager seat is agentic product work at startup velocity: (agentic warm-intro workflows), (autonomous podcast pipeline shown at Milan AI Week), and (A2A routing evaluators from the Google × Kaggle Agents Intensive). That is the technical bar founders inspect — and the bar he would set for Startups architects. Claude CMS remains a light production footnote, not another CMS-clone deep band.
- — solo build during Bolt.new's World's Largest Hackathon (May 2025), submitted to Y Combinator, now scaling through active enterprise pilots and CXO-level conversations.
- — Firecrawl scrape, Kimi-via-OpenRouter scripting, Gemini TTS on Azure render: the full pipeline shown live at Milan AI Week.
- — Google ADK hub-and-spoke architecture running six parallel country agents as FastAPI microservices for the Kaggle Agents Intensive capstone.
- Hiring bar already spans engineering, design, QA, and product — not only individual contributors — while scaling 4 → 22.
Manager motion · player-coach
- 01HireScreen for demonstrated build history over resume pattern-matching — proof over polish.
- 02ModelStay on strategic early-stage accounts with 0→1 receipts in hand
- 03CoachDeal coaching, eval playbooks, paired work with New Business AEs
- 04ScalePlaybooks so the team runs discovery without becoming a bottleneck
- EM → CIO at ELO2022–2024 player-coach run — the operating history behind the people pillar above.
- TagRamp · NewScriber · VisaPros0→1 receipts — full mechanism in the showcase above.
- Claude Code + Agent SkillsDaily Claude-led craft and Anthropic certifications — model the frontier habits, then scale through the team.
How agents get run
How I actually run agents
Claude-led for client delivery. Everything else is deliberate experiment: skills, hooks, harnesses, and loops that cut cost, tokens, and time-to-ship.
Skills from repeating patterns
When a workflow shows up more than once, I turn it into an agent skill with reusable instructions and triggers, rather than a one-off prompt.
Hooks, tools, and plugins for cost and speed
Most builds add hooks, tools, or plugins aimed at lower spend, tighter token use, and faster delivery. It is deterministic control in the places where prompts alone waste budget.
Harness and model fit per task
Claude Code for primary client build work; Antigravity and others when the task or token budget asks for it. This site's chat runs on-device Gemma 4 E2B (WebGPU + MediaPipe, model hosted on R2). NewScriber scripts with Kimi via OpenRouter and voices with Gemini TTS via Azure.
Technical inventory
Deep treatment: TagRamp, NewScriber, and VisaPros as the builder bar for the team — Claude CMS only as a light production footnote.
People leadership
- Hiring & structured onboarding
- 1-on-1 coaching (alumni in Norway · Canada · Germany)
- Pre-sales / technical GTM team leadership
- Player-coach bar for architecture quality
Builder bar
- Shipped LLM products (TagRamp · NewScriber · VisaPros)
- Claude in production for clients
- 0→1 founder credibility founders inspect
- MCP · Agent Skills · eval harnesses
Technical GTM
- Startup / high-growth company fluency
- Solution architecture for early buyers
- Executive & founder communication
- Ambiguous briefs → scoped delivery
Full-stack craft
- TypeScript / React / Node.js
- Python · Go
- Postgres
- Docker · CI/CD
Client & Delivery
- Enterprise problem identification
- Technical scoping
- Stakeholder alignment
- Delivery management
Analogous builds
Projects that map to this seat
A new LinkedIn for orgs that actually deal. TagRamp connects organizations from goals to meaningful business deals through AI-mediated intros. Built solo during the World's Largest Hackathon (May 2025) by Bolt.new and submitted to Y Combinator; now scaling with active enterprise pilots and CXO-level partnership discussions.
Agentic news scraper & editorial audio briefing network. Scrapes tech and business news via Firecrawl, curates & scripts dual-host dialogues with Kimi via OpenRouter, and renders high-fidelity multilingual voices via Gemini TTS on Azure.
Submitted as the capstone for the Google × Kaggle 5-Day AI Agents Intensive (2026). VisaPros is an agentic visa eligibility advisor utilizing a Google ADK agent-to-agent (A2A) hub-and-spoke architecture — parsing documents once and running six parallel country agents as FastAPI microservices to evaluate fit across destinations.
Coordinated ELO's dedicated engineering pod for Frontgo (Frontpayment's engineering arm) in Norway. The project itself was owned and driven by the client's dedicated project manager; Fauzul's role was to keep the three-engineer pod aligned — setting engineering practices, advising on the Vipps and financing integrations, unblocking communication and requirement gaps, leading hiring/HR, and running the weekly and monthly syncs — through to a clean handover to the in-house team.
Difficult questions
Candid answers on why manage rather than IC, SF relocation, and how hiring bars transfer from delivery pods to a GTM architect team.
You have not managed a formal pre-sales SA team — why this manager seat?
Because I have managed and grown the harder dual-mandate team: engineers who also had to earn customer trust and ship. EM → CIO at meant hiring, coaching, performance bars, and still sitting in the room on strategic accounts. The JD wants a player-coach who models great technical partnership — that is my default, not a stretch narrative. The ramp is Anthropic's SA-specific rituals, not first-time people leadership.
Will you relocate to San Francisco?
Yes. I relocate as soon as sponsorship processes and commit to SF hybrid office expectations and Pacific business hours from day one. Europe stays personal geography (sister in London, Nordic ties); this role's center of gravity is the SF startup ecosystem and I treat that as a real move.
How do you set the technical bar for architects advising startups on Claude?
I look past titles for demonstrated build history: has this person shipped something a founder actually adopted, and can they explain an eval they ran against real usage rather than a benchmark. , , and set that bar for me — full walkthrough is in the technical proof band above.
How do you partner with New Business AEs without becoming a bottleneck?
Clear playbooks for discovery, eval, and competitive bake-offs; personal ownership of a few strategic accounts; coaching the team on the rest. Same pattern I used scaling — I stayed on the hardest threads and built systems so the team could run the rest.
Beyond the role
What else Fauzul brings to the manager seat
Founder sales scar tissue, hiring systems, and public teaching — the JD's strong-candidate hints, translated into SA rituals.
Founder-led sales scar tissue
Priced deals, scoped SOWs, and sat with technical buyers as the person who also had to ship — firsthand read on what early-stage buyers evaluate.
Hiring and coaching infrastructure
Built onboarding, code-review and delivery playbooks while scaling 4 → 22 — raw material for SA hiring frameworks in a high-growth seat.
Public teaching for GTM presence
Developer workshops and a public learning series — comfortable building Anthropic's technical presence through content and rooms, not only closed-door POCs.
Start a conversation
Intentional about San Francisco — hybrid office-ready (~25%+), visa sponsorship required. Pacific hours from day one. Ready to hire and coach Startups architects while personally carrying strategic early-stage accounts.