Summary: Personal branding strategy for PhD-seeking AI researcher (MS→Direct PhD Fall 2027, full funding). Goal: position as credible, communicative, technically deep candidate for top advisors. Strategy: publish technical deep-dives, show reproducible work, engage with target lab researchers, maintain canonical portfolio.
Brand Positioning
Tagline: "Efficient AI at scale: sparse attention, world models, agentic systems"
Three Pillars:
| Pillar |
Evidence |
Platform |
| Technical Depth |
Paper breakdowns, math, benchmarks |
Blog (Substack/owned), GitHub |
| Research Vision |
Future directions, idea synthesis |
Blog, LinkedIn, X threads |
| Engineering Competence |
Code, reproduction, agent workflows |
GitHub, LinkedIn, blog posts |
Portfolio Assets (Must Have)
| Asset |
Status |
Deadline |
| Published blog: FlashAttention deep-dive |
Draft → Publish |
Aug 2026 |
| Published blog: LoRA deep-dive |
Draft → Publish |
Sep 2026 |
| Reproduction: DeepSeek-V4 / HRM / TRM |
In progress |
Oct 2026 |
| Project: Efficient ViT with DSA |
Planning → Active |
Dec 2026 |
| GitHub repo: ViT-DSA (500+ stars) |
Not started |
Mar 2027 |
| First-author paper submission |
Target: NeurIPS/ICML |
Oct 2026 / Feb 2027 |
| LinkedIn: 2K+ followers, technical Saves |
Building |
Ongoing |
| Substack: 500+ subscribers |
Building |
Ongoing |
Content Strategy (Recurring)
Monthly
- 1 long-form blog (2,000-4,000 words) on pillar topic
- 4-8 LinkedIn posts condensed from blog + comments
- 8-12 X threads on papers, ideas, engineering tips
Quarterly
- 1 paper reproduction with honest limitations
- 1 project update (code + benchmarks)
- Outreach: 5 personalized emails to target advisors
Networking Through Content
| Action |
Target |
Frequency |
| Comment on advisor papers |
Target PIs |
2×/week |
| Share paper breakdowns |
Lab members |
Per paper |
| Tag researchers in posts |
When relevant |
Natural |
| Reply to X discussions |
Technical threads |
Daily |
| DM with specific question |
After engagement |
1×/week |
Rule: Never cold-email without content reference. "I read your paper X, here's my breakdown Y."
Application Package Content
| Document |
Source |
Update Cycle |
| CV |
wiki/resume.md (auto) |
Monthly |
| Research Statement |
Blog posts + project pages |
Per application |
| SOP |
Personal narrative + content |
Per application |
| Portfolio URL |
ahsan.ai (owned domain) |
Always current |
| Code links |
GitHub pinned repos |
Per project |
Metrics to Track
| Metric |
Why |
Target |
| Blog: unique readers/post |
Reach |
1,000+ |
| LinkedIn: Saves/post |
Algorithm signal |
20+ |
| X: thread impressions |
Technical reach |
5,000+ |
| GitHub: stars (main repo) |
Code adoption |
500+ |
| Substack: subscribers |
Direct audience |
500+ |
| Citations/discussions |
Academic relevance |
5+/paper |
DON'Ts
- ❌ Don't chase viral (generic) content
- ❌ Don't post without canonical URL
- ❌ Don't exaggerate reproduction results
- ❌ Don't ignore limitations sections
- ❌ Don't auto-cross-post without adaptation
Related Concepts
Sources