Type / to search
YOU
your-team / main
Public
Contact

Add Chaoran Jiang to your team #2027

OpenMerged chaoran-jiang wants to mergemerged 9 commits into your-team:main from chaoran-jiang:new-grad-2027
CJ
chaoran-jiang opened this pull request · Aug 2026 author

What this PR does

Adds a Data Science senior at UW-Madison who works as a product owner. I write the spec, the roadmap and the acceptance criteria, then run AI agents as the engineering team that implements them. What comes out is software you can open, not slides about software.

Why now

I graduate in June 2027 and I am available full-time from then. Authorized to work in the U.S. without sponsorship.

How it was tested

Not in a sandbox. Three environments, three different ways to be wrong:

  • A buy-side research desk. Greenwoods Asset Management (景林资产), Shanghai, summer 2026. A global AI-video industry market-sizing report, delivered as a clean front-facing report plus an Excel backend where every figure links back to its source.
  • A chip company. xMEMS, Santa Clara, summer 2024. Finance datasets cleaned in Python and loaded into the accounting system, records reconciled in Excel, and the Tableau dashboard sitting on top of it.
  • A company I founded at fifteen. Ignite 3D Hologram Co., 2020 to 2023. A 3D hologram device, a national utility-model patent, a design patent, and two venture investors who sent me back to iterate.

Checklist

  • Every claim in this diff traces to a source
  • Work authorization confirmed, no sponsorship required
  • Available full-time from June 2027
  • Onboarding scheduled  blocked on reviewer
Where to start. The diff is all additions. Open projects/ if you care how I build, experience/ if you care where it has held up under someone else's standards.
chaoran-jiang added the labels new-grad, full-time, no-sponsorship-required Aug 2026
chaoran-jiang requested a review from you Aug 2026
All checks have passed 6 successful checks Show all checks
facts / verified Every claim on this page traces to a sourceChecking claims against sources 1.4s Details
work-authorization / status Authorized to work in the U.S. without sponsorship 0.9s Details
patents / registry 2 national patents: one utility model, one designQuerying patent registry 2.1s Details
ships-to-production / build 1 company founded, 2 internships, 3 systems builtBuilding 3.6s Details
availability / jun-2027 Full-time start on graduation, June 2027Reading calendar 0.6s Details
interests / human Vocal performance, fitness, badminton, pokerChecking for a pulse 0.4s Details
Review required At least 1 approving review is required to merge, and you are the requested reviewer. Changes approved 1 approving review by you.
This branch has no conflicts with the base branch Merging can be performed automatically. Notice period: none, he is a student.
You can also view command line instructions.
$ git fetch origin new-grad-2027 $ git checkout -b chaoran-jiang origin/new-grad-2027 $ git merge --no-ff chaoran-jiang -m "Add Chaoran Jiang to your team (#2027)" # or skip the ceremony: $ mail jiangchaoran28 [at] gmail [dot] com
Pull request successfully merged and closed The new-grad-2027 branch is now part of your-team:main.
YOU
Write Preview
9 commits on new-grad-2027 newest first · click any commit to expand
Commits in 2026
  • Produced the report under a buy-side mentor: multi-source triangulation of market bases held to one unified global scope, with penetration treated as an explicit forward-looking assumption rather than a borrowed number.
  • Shipped in two layers. A clean front-facing report for readers, and an Excel backend where every figure links back to the source it came from.
  • Acceptance bar: pick any number in the report, and I can show you the cell it traces to.
files experience/greenwoods.mdrole Equity Research Internreviewed by buy-side mentor
  • Covered AI-application equities across US, HK and A-share markets, synthesizing sell-side research from Goldman Sachs and CITIC-tier coverage into one view.
  • Built the sector dashboard on a market-data pipeline I wrote myself, so the coverage list refreshed without hand-copying prices.
files experience/greenwoods.mdtickers 33
  • Seven steps, from market selection to a thesis that names its own falsification signals to staged position sizing. If nothing can prove the thesis wrong, it is not a thesis.
  • Reviewed by a senior buy-side investor.
  • Ran an AI-assisted research workflow underneath it (multi-agent collection, cross-checking, data scraping) while owning every judgment and assumption personally.
files experience/greenwoods.mdsteps 7
  • A student uploads their DARS degree audit, picks a preference profile, and gets three or fewer complete semester schedules with an explainable trade-off behind every course and professor choice.
  • Unified three heterogeneous sources into one data model: the Madgrades grade-distribution API, UW Course Search & Enroll, and RateMyProfessor.
  • Python and FastAPI backend with a constraint-based recommendation engine.
  • Run as product owner with AI agents as the engineering team. I own the spec, the roadmap and the acceptance criteria; implementation is delegated to multi-agent workflows.
files projects/course-planner.pyrole Founder & Product Ownerstatus in development
  • End-to-end daily pipeline in Python: an interpretable stock-selection model on price, volatility, drawdown and volume features, with logistic regression for next-day move probabilities.
  • Cross-sectional ranking picks a concentrated 3 to 5 stock portfolio out of a roughly 60-ticker high-liquidity universe, under realistic constraints (integer shares, cash buffer, sizing rules).
  • Data validation and freshness checks to prevent look-ahead bias. Evaluated in paper trading with drawdown-focused risk control.
  • LLMs translate investment constraints into rule-based logic and stress-test features, which keeps the whole system explainable.
files projects/quant-system.pyuniverse ~60 tickersportfolio 3 to 5 names
  • A standing team of agents (researcher, architect, product manager, designer) that shares one context and gets routed by task type, rather than being spun up from scratch every session.
  • Persistent memory built as a knowledge graph linking projects, skills and decisions.
  • Scheduled automations distill each week's corrections into upgrades to the system itself. The mistakes become next week's rules.
files projects/aios.yamlagents 4 standing roles
Commits in 2024
  • Cleaned and transformed finance datasets in Python and loaded them into the accounting system.
  • Maintained and reconciled accounting records in Excel.
  • Contributed to a Tableau dashboard visualizing operations and sales performance.
  • Assisted SEO keyword analysis in SEMrush that informed website improvements.
files experience/xmems.mdrole Data Analyst Intern
Commits in 2023
  • B.A. in Data Science, Sep 2023 to Jun 2027 (expected).
  • Coursework that shows up in the projects: Data Science Modeling I & II (STAT 240 / 340), Data Science Programming (COMP SCI 220), Elementary Matrix & Linear Algebra, Principles of Micro- and Macroeconomics.
files education/uw-madison.mdgraduating Jun 2027
Commits in 2020
  • Designed a 3D hologram device.
  • Secured a national utility-model patent and a design patent.
  • Pitched the business plan to two venture investors and iterated on their feedback.
  • Ran until Jun 2023, which is when the runtime moved to Madison.
files experience/ignite.mdrole Founderpatents 2
facts / verifiedSucceeded in 1.4s
Run facts/verify@v1 scanning 9 files education/uw-madison.md ....... ok experience/greenwoods.md ...... ok experience/xmems.md ........... ok experience/ignite.md .......... ok projects/course-planner.py .... ok projects/quant-system.py ...... ok projects/aios.yaml ............ ok skills.json ................... ok interests.md .................. ok no unsourced claims Done in 1.4s
Run authorization/check@v2 status ........... authorized sponsorship ...... not required verbatim ......... Authorized to work in the U.S. without sponsorship Done in 0.9s
Run patents/lookup@v1 registrant ....... Ignite 3D Hologram Co. utility model .... granted national design ........... granted national filed at age 15 Done in 2.1s
Run build --all ignite-3d-hologram ..... company, 2020 to 2023 course-planner ......... in development quant-system ........... paper trading aios ................... running daily 4 artifacts, 0 slide decks Done in 3.6s
Run availability --resolve graduation ....... Jun 2027 start ............ full-time, on graduation base ............. Madison, WI / Hangzhou, CN Done in 0.6s
Run human --detect vocal performance .... 6 yrs, top-4 in a university competition fitness .............. 5 yrs, data-driven badminton ............ yes poker ................ only with an edge pulse detected Done in 0.4s
9 changed files with 124 additions and 0 deletions
education/uw-madison.md +10
@@@@ @@ -0,0 +1,10 @@ education/uw-madison.md
1+# University of Wisconsin-Madison
2+
3+**B.A. in Data Science** · Sep 2023 - Jun 2027 (expected)
4+Madison, Wisconsin
5+
6+## Relevant coursework
7+- Data Science Modeling I & II (STAT 240 / 340)
8+- Data Science Programming (COMP SCI 220)
9+- Elementary Matrix & Linear Algebra
10+- Principles of Micro- and Macroeconomics
experience/greenwoods.md +20
@@@@ @@ -0,0 +1,20 @@ experience/greenwoods.md
1+# Equity Research Intern
2+**Greenwoods Asset Management (景林资产)** · Shanghai · Jun 2026 - Aug 2026
3+
4+## Market sizing
5+- Produced a global AI-video industry market-sizing report under a buy-side mentor:
6+ multi-source triangulation of market bases held to one unified global scope, with
7+ penetration treated as an explicit forward-looking assumption
8+- Delivered a clean front-facing report plus a fully traceable Excel backend where
9+ every figure links back to its source
YOU
you commented on line 9
Every figure? That is a strong claim.
CJ
chaoran-jiang replied
Pick a number in the report and I will open the cell it came from. That was the acceptance bar I set with my mentor, and it is why the delivery is two layers: a clean report for readers, an Excel backend for anyone who wants to audit it.
10+
11+## Coverage
12+- Covered AI-application equities across US, HK and A-share markets, synthesizing
13+ sell-side research (Goldman Sachs, CITIC-tier coverage)
14+- Built a 33-ticker sector dashboard on a self-built market-data pipeline
15+
16+## Process
17+- Co-developed a 7-step investment decision framework (market selection, thesis with
18+ falsification signals, staged position sizing), reviewed by a senior buy-side investor
19+- Ran an AI-assisted research workflow (multi-agent collection, cross-checking, data
20+ scraping) while owning every judgment and assumption personally
experience/xmems.md +8
@@@@ @@ -0,0 +1,8 @@ experience/xmems.md
1+# Data Analyst Intern
2+**xMEMS** · Santa Clara, CA · May 2024 - Aug 2024
3+
4+- Cleaned and transformed finance datasets in Python and loaded them into the
5+ accounting system
6+- Maintained and reconciled accounting records in Excel
7+- Contributed to a Tableau dashboard visualizing operations and sales performance
8+- Assisted SEO keyword analysis (SEMrush) that informed website improvements
experience/ignite.md +8
@@@@ @@ -0,0 +1,8 @@ experience/ignite.md
1+# Founder
2+**Ignite 3D Hologram Co.** · Aug 2020 - Jun 2023
3+
4+> Founded at fifteen.
YOU
you commented on line 4
wait, at fifteen?
CJ
chaoran-jiang replied
Fifteen. Both patents are filed under my name. The pitch decks were much worse than the device, which is roughly what the two venture investors told me, and that is where the iterating started.
5+
6+- Designed a 3D hologram device
7+- Secured a national utility-model patent and a design patent
8+- Pitched the business plan to two venture investors and iterated on their feedback
projects/course-planner.py +22
@@@@ @@ -0,0 +1,22 @@ projects/course-planner.py
1+"""UW-Madison Course Planner. In development."""
2+
3+# Founder & Product Owner.
4+# product owner: I own spec, roadmap, acceptance.
5+# engineering team: AI agents running multi-agent workflows.
YOU
you commented on line 5
So who actually wrote the implementation?
CJ
chaoran-jiang replied
The agents did. I wrote the spec, the roadmap and the acceptance criteria, and I rejected whatever failed them. That is the job I want: deciding what gets built and what counts as done, with a team that can move as fast as I can specify.
6+
7+from sources import madgrades, course_search_enroll, rate_my_professor
8+
9+SOURCES = (madgrades, course_search_enroll, rate_my_professor) # 3 feeds, 1 data model
10+
11+def solve(dars_audit: DegreeAudit, prefs: Preferences) -> list[Schedule]:
12+ """A student's DARS degree audit in. Three or fewer schedules out."""
13+ remaining = dars_audit.unmet_requirements()
14+ catalog = unify(SOURCES) # grade distributions + offerings + professor signal
15+ candidates = constraint_engine.search(remaining, catalog, prefs)
16+
17+ # conflict-free is table stakes. the reason attached to each pick is the product.
18+ ranked = rank(candidates, explain=True)
19+ return ranked[:3]
20+
21+# stack: Python / FastAPI, constraint-based recommendation engine
22+# status: in development. no user numbers on this page until it ships.
projects/quant-system.py +18
@@@@ @@ -0,0 +1,18 @@ projects/quant-system.py
1+"""AI-assisted quantitative trading system. US equities, daily frequency."""
2+
3+UNIVERSE = 60 # high-liquidity tickers
4+PORTFOLIO = (3, 5) # concentrated, not diversified into noise
5+FEATURES = ["price", "volatility", "drawdown", "volume"]
6+
7+def daily_pipeline(as_of: date) -> Orders:
8+ prices = load_prices(as_of)
9+ assert no_look_ahead(prices, as_of) # data validation + freshness checks
10+
11+ p_up = logistic_regression.predict_proba(features(prices, FEATURES))
12+ picks = cross_sectional_rank(p_up)[:PORTFOLIO[1]]
13+
14+ return size(picks, integer_shares=True, cash_buffer=True, rules=SIZING_RULES)
15+
16+# LLMs translate investment constraints into rule-based logic and stress-test features.
17+# the system stays explainable: a position I cannot explain does not get taken.
18+# evaluated in paper trading, risk control centred on drawdown.
projects/aios.yaml +20
@@@@ @@ -0,0 +1,20 @@ projects/aios.yaml
1+# AIOS: a personal AI operating system
2+
3+team: # standing, not spun up per task
4+ - researcher
5+ - architect
6+ - product_manager
7+ - designer
8+context: shared # one context, routed by task type
9+
10+memory:
11+ kind: knowledge_graph
12+ links: [projects, skills, decisions]
13+ survives: session_end
14+
15+automation:
16+ weekly:
17+ - distill: corrections # what went wrong this week
18+ into: system_upgrades # becomes next week's rules
19+
20+operator: chaoran-jiang
skills.json +12
@@@@ @@ -0,0 +1,12 @@ skills.json
1+{
2+ "languages": ["Python", "SQL"],
3+ "python": ["pandas", "NumPy", "scikit-learn"],
4+ "analysis": ["Tableau", "Advanced Excel"],
5+ "llm_workflows": [
6+ "multi-agent orchestration",
7+ "prompt engineering",
8+ "AI-assisted research pipelines"
9+ ],
10+ "operating_mode": "product owner: spec, roadmap, acceptance",
11+ "authorization": "Authorized to work in the U.S. without sponsorship"
12+}
interests.md +6
@@@@ @@ -0,0 +1,6 @@ interests.md
1+# Outside the diff
2+
3+- Vocal performance: 6 years of training, top-4 in a university competition
4+- Data-driven fitness training: 5 years
5+- Badminton
6+- Poker: I only play the hands where I have an edge
Merged

Merged. Welcome aboard.

9 commits from chaoran-jiang are now part of your-team:main.

Chaoran Jiang

Madison, WI / Hangzhou, China  |  jiangchaoran28 [at] gmail [dot] com  |  linkedin.com/in/jiangchaoran28
Authorized to work in the U.S. without sponsorship

Education

University of Wisconsin-MadisonSep 2023 - Jun 2027 (expected)
B.A. in Data Science
  • Relevant coursework: Data Science Modeling I & II (STAT 240 / 340), Data Science Programming (COMP SCI 220), Elementary Matrix & Linear Algebra, Principles of Micro- and Macroeconomics

Professional experience

Equity Research Intern, Greenwoods Asset Management (景林资产)Jun 2026 - Aug 2026
Shanghai, China
  • Produced a global AI-video industry market-sizing report under a buy-side mentor: multi-source triangulation of market bases with unified global scope, penetration treated as explicit forward-looking assumptions; delivered as a clean front-facing report plus a fully traceable Excel backend where every figure links to its source
  • Covered AI-application equities across US, HK and A-share markets: synthesized sell-side research (Goldman Sachs, CITIC-tier coverage) and built a 33-ticker sector dashboard with a self-built market-data pipeline
  • Co-developed a 7-step investment decision framework (market selection, thesis with falsification signals, staged position sizing), reviewed by a senior buy-side investor
  • Ran an AI-assisted research workflow (multi-agent information collection, cross-checking, data scraping) while owning every judgment and assumption personally
Data Analyst Intern, xMEMSMay 2024 - Aug 2024
Santa Clara, CA
  • Cleaned and transformed finance datasets in Python and loaded them into the accounting system
  • Maintained and reconciled accounting records in Excel
  • Contributed to a Tableau dashboard visualizing operations and sales performance
  • Assisted SEO keyword analysis (SEMrush) informing website improvements
Founder, Ignite 3D Hologram Co.Aug 2020 - Jun 2023
Founded at fifteen
  • Designed a 3D hologram device; secured a national utility-model patent and a design patent
  • Pitched the business plan to two venture investors and iterated on their feedback

Projects

UW-Madison Course Planner (in development)Founder & Product Owner
  • Building a web product that turns a student's DARS degree audit into three or fewer ranked, conflict-free semester schedules, with explainable trade-offs behind every course and professor choice
  • Unified three heterogeneous data sources (Madgrades grade-distribution API, UW Course Search & Enroll, RateMyProfessor) into one data model; Python/FastAPI backend with a constraint-based recommendation engine
  • Run as product owner with AI agents as the engineering team: I own spec, roadmap and acceptance; implementation is delegated to multi-agent workflows
AI-Assisted Quantitative Trading SystemUS equities, daily frequency
  • Designed an end-to-end daily pipeline in Python: interpretable stock-selection model on price, volatility, drawdown and volume features; logistic regression for next-day move probabilities
  • Cross-sectional ranking selects a concentrated 3 to 5 stock portfolio from a roughly 60-ticker high-liquidity universe, under realistic constraints (integer shares, cash buffer, sizing rules)
  • Built data validation and freshness checks to prevent look-ahead bias; evaluated via paper trading with drawdown-focused risk control
  • Used LLMs to translate investment constraints into rule-based logic and stress-test features, keeping the whole system explainable
AIOS, Personal AI Operating Systemin daily use
  • Designed and operate a personal AI operating system: a standing multi-agent team (researcher, architect, product manager, designer) sharing one context, routed by task type
  • Built persistent memory with a knowledge graph linking projects, skills and decisions, plus scheduled automations that distill weekly corrections into system upgrades

Skills

Python (pandas, NumPy, scikit-learn), SQL, Tableau, Advanced Excel; LLM-powered workflows: multi-agent orchestration, prompt engineering, AI-assisted research pipelines

Interests

Vocal performance (6 yrs training, top-4 in university competition) · data-driven fitness training (5 yrs) · badminton · poker

Branch new-grad-2027 was not deleted.

Every line in this diff is a +. The deletions happened before the commit.
First commit to your-team:main lands June 2027.

9 of 9 files reviewed. Nothing left to check. The merge button is waiting in the Conversation tab.

Finish your review
Changes requested.

Fair. That path is well tested.

Two venture investors requested changes on the Ignite business plan in 2020, and he iterated on their feedback. A senior buy-side investor reviewed the 7-step investment framework this summer. A buy-side mentor set the bar that every figure in the market-sizing report had to trace back to its source, which is why it shipped with an Excel backend attached.

Revision is the normal path here, not the exception. Leave the note and keep reading, or merge and send the note afterwards.

9 of 9 files still unviewed
Done