FLAGSHIP CERTTier-1
AI, Data & Technology
certifications
Flagship AI and data-science credentials across Tier-1 issuers (IBM, Google, Cisco, DataCamp, Ateneo de Davao).
proof/ · show your work
Proof. Public or self-owned only.
A gallery of public, NDA-clean artifacts — records, charts, co-signed certificates — that back up the claims on/experience,/certifications, and /now. Thirty-one artifacts on disk, twenty-two surfaced below. Updated when something new ships; never deleted when something expires.
v5.2 · 2026-07-10
P.2 · Flagship credentials
Tier-1 certs. Co-signed by name.
Three flagship credentials already curated on/certifications plus one standalone Tier-1 from Bitget. Each card carries the issuer, date, and the photo.
FLAGSHIP CERTTier-1
Finance, Trading & Economics
certifications
Tier-1 finance credentials: STA flagship + Goldman / JPMorganChase brand-name work + DataCamp finance stack.
FLAGSHIP CERTTier-1
Events, Hackathons & Distinguishing Flagships
certifications
Distinguishing Tier-1 flagships that anchor a memorable profile: NASA Space Apps, Meta BIDA, university teaching, and Philippine civil-service eligibility.
FLAGSHIP CERTTier-1
Blockchain4Youth · B4Y-2026-000701
AI · Data · Technology
Bitget Blockchain4Youth · Tier-1 distinction · 2026.
P.3 · Shipped code
Things I can import.
Four repositories. Each is MIT-licensed,offline-runnable, and dependency-light. Each one is a small, auditable artifact that demonstrates the eval-first methodology in code rather than prose.
MCP server
mcp-backtest-server
IP-clean backtest tools over the Model Context Protocol. Three tools, four textbook strategies, zero market data.
100% offline · 4 strategies · 0 lines of proprietary data
$ uv run server.py
# FastMCP server listening on stdio
{"tools":["get_ohlcv","run_backtest","compute_metrics"]}
smoke · 262 bars returned
sma_cross sharpe=0.471 cagr=0.079 max_dd=-0.21
buy_and_hold sharpe=-0.490
PASS · deterministic from symbol seed
Python package
qfin-rag-harness
Citation-grounded retrieval harness over a curated corpus of 16 canonical q-fin papers. Pure-Python TF-IDF cosine, no LLM call.
16 papers · 0 API keys · 0 LLM dependency
$ python -m qfin_rag "How do I correct factor significance for data mining?"
# Citations for: How do I correct ...
## Harvey (2016) — And the Cross-Section ...
*Review of Financial Studies, 2016* (relevance: 0.173)
query → 16-paper curated corpus
top-3 by TF-IDF cosine
citation-grounded prompt template
NO network calls
Eval harness
numerical-faithfulness-eval
Verify LLM numerical claims against deterministic fixtures. Catches the most common LLM failure mode — wrong numbers — in under one second.
10 demo claims · 5 pass · 5 fail by design
$ python -m eval_harness
[FAIL] btc_2024_sma_cross_20_50 / sharpe: claimed=5.0,
truth=0.4707, diff=4.5293 > tol 0.05
=== summary ===
5/10 passed (50.0%)
fixtures computed deterministically from seeded backtest
tolerances per metric (Sharpe atol=0.05, CAGR=0.02)
per-claim reason + diff vs tolerance
exit code = number of failed claims
MCP server
eval-mcp-server
Slop-evaluation gate exposed as an MCP tool. Tools / Resources / Prompts — 20/20 conformance, 100% round-trip parity.
MCP conformance 20/20 · round-trip parity 100% · primitives 3/3
$ uv run eval_mcp_server.py
# MCP server with the slop-evaluation gate
{"primitives":["tools","resources","prompts"],
"tools":["evaluate_text","score_metrics","emit_report"]}
20/20 conformance cases pass
round-trip parity across all 3 primitives
Tools · Resources · Prompts all live
PASS · slop score reproducible per call
P.4 · Research artifacts
9 quant · 6 AI · 10 solutions. All public-data reproducible.
Compact list rendering (AQR-Insights style). Each row links to the project or solution detail page; metrics come from the project frontmatter, not invented.
Quant projects — 9
- QUANTMultiple Testing & the Deflated Sharpe Ratio1.14Best-of-160 in-sample SharpeSep 2025
Best-of-160 BTC rule: IS Sharpe 1.14 is only 1.24× the pure-noise expectation — DSR = 0.70 (fail).
- QUANTCross-Sectional Momentum (18 coins)0.91In-sample SharpeOct 2025
IS Sharpe 0.91 → OOS −0.03 — an honest decay; bootstrap CI straddles zero.
- QUANTTime-Series Momentum + Vol Targeting0.27Sharpe (raw)Dec 2025
Vol-targeting lifts Sharpe 0.27 → 0.39 and halves max DD (−62% → −30%).
- QUANTThe Variance Risk Premium (VIX vs Realized)85%Years VRP > 0Jan 2026
Implied > realized 85% of 36 yrs; predicts returns, Newey-West t = +6.5.
- QUANTPairs Trading via Cointegration (BTC / ETH)−2.11ADF test statisticFeb 2026
ADF −2.11 (not cointegrated), half-life 208 d — the fade loses, as the test predicts.
- QUANTCrypto Funding-Carry+11.9%Annualized fundingApr 2026
Funding +11.9% annualized premium; fade IS 1.11 → OOS −0.05 (decayed post-2023).
- QUANTMacro / Volatility-Regime Overlay0.64Sharpe (base)May 2026
Vol-managed exposure: Sharpe 0.64 → 0.67, max DD −58% → −42%.
- QUANTBacktest Engine + Cost Model0.20Sharpe (gross)May 2026
High-turnover signal wins gross (0.20 > 0.13) but loses net of cost — break-even 20 bps.
- QUANTLook-Ahead Bias Audit (the shift test)0.59Clean SharpeJun 2026
A leak inflates Sharpe 0.59 → 5.07; the shift test exposes it as 88% phantom.
AI projects — 6
- AIrag-recall0.886recall@3Nov 2025
RAG service that proves its own retrieval — recall@3 = 0.886, MRR@3 = 0.805 with offline stdlib TF-IDF retriever.
- AItoolcall-agent100%Tool / arg correctnessDec 2025
ReAct-style tool-calling agent with OTel traces, fault injection, and 100% tool/arg correctness.
- AIjudge-harness0.58Cohen's κ (vs human)Jan 2026
LLM-as-judge pipeline validated against human raters — Cohen's κ = 0.58 with bootstrap CI and position-bias measured.
- AIeval-mcp-server20 / 20MCP conformanceFeb 2026
MCP server exposing the slop-evaluation gate over Tools, Resources, and Prompts — 20/20 conformance, 100% round-trip parity.
- AIreflect-revise127.5 → 14.0Mean SLOP scoreMar 2026
Reflection-loop agent — mean SLOP 127.5 → 14.0 across drafts, 3/4 improved, 1/4 honest no-progress halt.
- AIslop-scanner81 → 3Real-draft improvementMar 2026
13-metric literature-grounded AI-output quality gate — drove a real draft from HEAVY (81) to CLEAN (3).
Solutions — 10 case studies
- AI11-Agent Eval-First Research Platform~27,500 words of role-scoped agent charters
A small systematic-trading desk needed an AI workflow that could keep pace with the research pipeline *without* shipping hallucinated or unverifiable analysis. The naive path — a single agent with one prompt — produced plausible but unfalsifiable outputs. The desk needed something closer to a research organization than a chatbot.
- AI7-Agent Venture Incubation Pipeline31 decision-grade artifacts (charters, briefs, ledger entries) shipped
An operator-led venture incubation arm needed a repeatable way to triage, brief, and ledger new business ideas — without one human bottleneck, and without the agent that proposes a decision being the same one that approves it.
- AI8-Agent Editorial Production Pipeline (SLOP ↓ 96%)SLOP index dropped from 81 → 3 on the working corpus
A high-volume editorial workflow was generating content with a measurable "slop index" — generic, templated, easily-detected text. Quality gate was after-the-fact and manual. Production scaled faster than the editorial team could review.
- QUANT9-Project Public-Data Quant Research Library9 projects, each with methodology declared before results
Most online quant research demos are not reproducible: closed datasets, undisclosed parameters, un-reported multiple-testing bias, and no OOS discipline. A hiring-grade research portfolio needs every one of those addressed explicitly.
- QUANTDeflated Sharpe Ratio as a Built-in Pipeline Gatescipy-free numpy implementation (deploys anywhere)
Naive Sharpe ratios ignore the search effort — if you try 100 variants of an idea, the best-looking one will overstate the true edge. The desk needed this multiple-testing correction built into the validation pipeline, not stapled on at the end.
- AIEval MCP Server — 31 Gates as First-Class ToolsMCP-compliant server (Claude Agent SDK / Cursor / etc. can connect)
LLM eval harnesses live in code or in spreadsheets — neither is a clean integration target for multi-agent pipelines. A multi-agent system needs the eval gates exposed as **tools**, not as Python imports.
- QUANTCrypto Statistical-Arbitrage Pipeline with Funding-CarryFunding-carry integrated into P&L (not an afterthought)
Crypto perps run funding payments every 8h. A naive long-short book ignores funding carry and bleeds slowly when the spread is inverted. A working stat-arb needs the funding cost added back to P&L *before* sizing.
- QUANTTransaction-Cost-Aware Backtest EngineSpread + slippage + latency modeled per asset class
Backtests that ignore transaction costs, slippage, and latency are the most common source of overfit research. A backtest engine needs all three treated as first-class inputs, not as after-the-fact deductions.
- QUANTLook-Ahead-Bias Audit SuitePoint-in-time dataset verification
Look-ahead bias is silent: a backtest that uses future data looks great until you deploy it. A serious research portfolio needs an **explicit audit suite** — a checklist of failure modes with mechanical tests.
- QUANTPublic Finance Curriculum (CTA-Track, Self-Directed)STA Tier-1 CTA certified (program completed Dec 2025)
The Philippines has limited access to rigorous CTA-grade technical analysis training. Most curriculum is either imported (US/UK, expensive) or shallow (TA-by-rote). There's a gap for affordable, rigorous, evidence-based technical analysis education.
P.5 · Public track record
Lectures, charts, and feeds.
Two public lectures for USeP EGE 313, one public TradingView chart, and three Atom feeds (portfolio updates · project updates · solution updates).
Operator-discretion chart
Atom feeds — 3 streams
- /feed.xml · portfolio updates
- /feed-projects.xml · project updates
- /feed-solutions.xml · solution updates
P.6 · What's not here
NDA-positive disclosure.
What's not here: photos of the 19V Capital office. Screenshots of proprietary strategy outputs. Calendar entries with client names. The 19V Capital role is a closed past contract (03/2026 – 06/2026), and everything on this page is public-data or self-owned. If the proof you need isn't on this page, email me — if it's public, I'll send it within 24 hours.
Want a specific artifact?
Email me with the proof you need. If it's public, NDA-clean, and I have it on disk, I'll send it within 24 hours.