Ishaan Gupta

21, Building @Supermemory, EF Fall'26
prev. GeminiCLI @google, GSoC'25 @Alaska, NSUT'26

[Ishaan Gupta @_ishaanxgupta]

it's been two months since i joined @supermemory
with @DhravyaShah

building the best memory layer for AI agents alongside an incredible team.

[Ishaan Gupta @_ishaanxgupta]

Earlier this year, I got the chance to contribute to @geminicli, was a really fun experience working with the team, especially @TommSciortino, and seeing how fast the project moved.

Now that it has transitioned toward @antigravity, it feels special to see some of that work live on in the next chapter.

Some of the contributions I worked on included:
- prompting users to run /terminal-setup
- policy-related command improvements
- onboarding flow fixes
- hook system migrations

Open source is wild because a PR you make in January can quietly become part of something much bigger later.

My PRs:
https://t.co/c1F4bztQOx

[Ishaan Gupta @_ishaanxgupta]

@attharrva15 Engine😂😂

[Ishaan Gupta @_ishaanxgupta]

@GoogleDeepMind PR: https://t.co/5cWKgFaJ2j

Official benchmarked results: https://t.co/WqhWVMOFKZ

[Ishaan Gupta @_ishaanxgupta]

My PR was merged into @GoogleDeepMind's official Physics-IQ benchmark repository.
It's incredible how quickly i2v models have evolved.

7 months ago, I benchmarked CogVideoX-5B-i2v on Physics-IQ. It scored 32.3%.

At the time, that seemed pretty modest. The pace of progress in video generation is insane.

[Dhravya Shah @DhravyaShah]

something super cool is launching on supermemory's discord very soon...

Join our discord to know what it is 👀

[Ishaan Gupta @_ishaanxgupta]

@rrodrigu3z yess, give it a try!

[Ishaan Gupta @_ishaanxgupta]

best plugin ecosystem for AI coding agents.
Claude Code, Codex, Cursor, OpenCode, OpenClaw, Hermes...

One memory. Every coding agent.

[Ishaan Gupta @_ishaanxgupta]

Masterclass on our memory system!

[supermemory @supermemory]

Episode 1 of memory 101: the anatomy of memory.🧠

real memory isn't one thing. it's five:

> facts
> profiles
> relationships
> temporality
> forgetting

"AI memory" products build one of these primitives and stop. Real memory is when all five compose.

Next, stay tuned for RAG ↓

[Ishaan Gupta @_ishaanxgupta]

if you wanna deep dive into SMFS, read the technical report here:
https://t.co/vo62GqHjcC

[supermemory @supermemory]

Quick explainer on SMFS and why we think agent memory has been measured wrong ↓

everyone optimizes for retrieval speed. but the real cost is how many lookups it takes to answer, often ten tool calls a question, each one burning tokens.

SMFS goes after the count. it mounts your supermemory as a real filesystem:

→ the agent walks it with ls, cat, grep
→ except grep is semantic, scoped to the folder it's standing in
→ so it narrows by path instead of re-searching the whole index

on xAFS, our open benchmark:

→ 53% fewer tokens per answer, near-identical accuracy
→ 6.7 tool calls to an answer, down from 10.2
→ at 10k files: 13M tokens the old way vs 2.4M, the gap grows with scale

cut the number of retrievals and everything downstream gets cheaper.

<speed was never the metric>

[Ishaan Gupta @_ishaanxgupta]

currently on the max plan, worth every dollar !

[Dhravya Shah @DhravyaShah]

Introducing the $100 max plan on @supermemory cloud

- 7x limit from pro plan and 10x cheaper than self-hosting same amount of usage
- Granola, Gmail and other connectors
- Most API features enabled - Best for startups, internal team and heavy personal use!

[Ishaan Gupta @_ishaanxgupta]

Flying to Banglore sponsored by @join_ef
Super excited😁

[Ishaan Gupta @_ishaanxgupta]

@vedx09 @join_ef 🔥

[Ishaan Gupta @_ishaanxgupta]

Flying to Banglore sponsored by @join_ef
Super excited😁

[Ishaan Gupta @_ishaanxgupta]

@vedantxn @join_ef Did not get your dm

[Ishaan Gupta @_ishaanxgupta]

best plugin ecosystem for AI coding agents.
Claude Code, Codex, Cursor, OpenCode, OpenClaw, Hermes...

One memory. Every coding agent.

[Ishaan Gupta @_ishaanxgupta]

See every space across all your connected plugins in one view and jump between them instantly.

[supermemory @supermemory]

Your agent only knows what you've uploaded. Docs change, emails land, commits ship, none of it reaches your agent.

Connectors fix that. Plug in your sources once. Supermemory syncs the rest.

How it works:

→ You authenticate once. Setup, extraction, and indexing run automatically from there.
→ Sync is real-time. When something changes on the other end, a webhook reaches Supermemory within seconds.
→ Only the content that actually changed gets re-processed, never your entire drive.

The result is simple: your agents always work from the latest version of your knowledge, with nothing to maintain.

Stop importing. Start connecting.

[Ishaan Gupta @_ishaanxgupta]

The supermemory chat just got a serious upgrade.

• Grok 4.3 alongside GPT, Claude &amp; Gemini
• Attach files &amp; images
• Thinking mode for deeper reasoning

One chat. Backed by @supermemory.

[Ishaan Gupta @_ishaanxgupta]

The supermemory chat just got a serious upgrade.

• Grok 4.3 alongside GPT, Claude &amp; Gemini
• Attach files &amp; images
• Thinking mode for deeper reasoning

One chat. Backed by @supermemory.