- Following 84
- Followers 39.1K
- Engagement 5.56%
- Avg likes 1.5K
- Avg comments 701
About Shirin Khosravi Jam
Shirin Khosravi Jam (@jam.with.ai) is a Science, Technology & Engineering creator on Instagram, also filed under Technology. The account has 39,132 followers and 89 published posts. Recent posts average 1,514 likes and 701 comments, an engagement rate of 5.56%.
With 39.1K followers, Shirin Khosravi Jam is a micro creator by the common 10,000 to 100,000 definition. Set against 84 accounts followed, the audience works out to about 466 followers per account Shirin Khosravi Jam follows. Shirin Khosravi Jam carries Instagram's verified badge and uses an Instagram professional account. Flinque files Shirin Khosravi Jam under Science, Technology & Engineering, with Technology as a secondary category.
The Instagram bio for Shirin Khosravi Jam runs to 20 words across 4 lines. Emoji sit alongside the words. The latest activity Flinque has on record for Shirin Khosravi Jam dates from October 2025.
Story Highlights
Unlocking shows the title and cover image of every story highlight this creator keeps pinned to the profile.
89 Posts
Of the 10 most recent posts shown for Shirin Khosravi Jam, 6 are videos and 4 are carousels. That mix leans toward videos. 89 published posts give Shirin Khosravi Jam a moderate archive on Instagram. Against the audience, that is roughly 440 followers for each post published. Shirin Khosravi Jam also keeps 1 story highlight on the profile. All 10 carry captions, averaging 66 words each. The captions are written mainly in English. Words that recur across them include comment, data, hope, it’s and llms.
It’s important that people understand the benefits of generalised LLM and specialised SLMs!
LLMs are amazing.
But SLMs are how a lot of real products will ship: cheaper, faster, more private, and often “good enough” for the task.
Want a part 2 on when to choose SLM vs LLM?
The 10% ask → design → adapt under constraints.
Here’s the exact template + my 8 kickoff questions
Comment “system” for my 1-page diagram + example RAG baseline.
Save this for your next interview.
SYSTEM Kickoff Questions
1. What are the SLOs? (latency, cost)
2. Expected peak and growth?
3. Cost envelope (per request / monthly)?
4. Privacy/PII/region constraints?
5. Data freshness needs (batch vs streaming)?
6. How is quality measured in prod? (nDCG / precision@k / CTR)
7. Preferred fallback/degradation if constraints collide?
8. Which existing systems must we integrate or reuse?
I’ve shipped Data Science, MLOps, RAG systems, and AI agents
𝗠𝘆 𝗵𝗼𝗻𝗲𝘀𝘁 𝘁𝗮𝗸𝗲:
Pick 1-2 books.
Build something small.
Come back when you hit problems.
PS: There are many books out there, but these are the list of must read from my end. For learning LLMs as well there are multiple resources, so didnt include.
(what actually matters - beyond scikit-learn, ML libraries, etc.)
uv
→ fast, reproducible Python environments
ruff
→ linting + formatting without silent tech debt
pytest
→ tests for data, features, and ML logic
pydantic
→ strict schemas and data contracts
FastAPI
→ serve models as real APIs
Docker
→ reproducible, portable execution
Makefile
→ one-command, repeatable workflows
DVC
→ version datasets and models like code
Great Expectations
→ catch bad data before models break
Polars
→ fast, modern dataframes at scale
Airflow / Dagster
→ reliable production pipelines
MLflow / Weights & Biases
→ track experiments and artifacts
Postgres
→ retrieval and storage fundamentals
Grafana
→ monitor latency, errors, and business metrics
Git (properly)
→ collaboration without chaos
It’s exactly 1.5 years since I started working on them in production.
I’m just reflecting some main gaps I saw between a data science professional and the new me who deployed AI agent and RAG in production.
I really hope it’s helpful 😇
Also I must say that this certificate was big boost in my learning and my career!
Hope it helps you 😇
Two things separate strong AI/ML engineers from the rest. No hype. Just fundamentals.
I hope it is helpful😇
Unsloth AI fine-tunes a LLM with just 3GB VRAM.
It makes this possible with QLoRA + dynamic 4-bit quants.
Here’s what you can actually do:
→ Inject domain-specific knowledge into any model
→ Customize tone, personality, response style
→ Train on your own data for specialized tasks
What makes this beginner-friendly:
↳ Pre-made Colab notebooks (zero setup)
↳ Synthetic dataset generator included
↳ Works on Colab, Kaggle, or local GPU
Start with Llama 3.1 8B Instruct.
Use QLoRA for 4-bit training.
Get results in hours, not days.
Full hands-on guide covers everything:
- Dataset creation
- Hyperparameter tuning
- Model evaluation
- Deployment to Ollama/vLLM
I hope it helps:)
Many RAG fail because OCR fails!
If you’re building GenAI apps that touch documents, OCR is foundational.
These 6 open-source libraries are leading the charge in 2026👇
Know them well:)
Start treating OCR like a first-class citizen in your pipeline.
It’s not just about LLMs anymore. It’s about what you feed them.
It’s all about how well you pre-process your data!
You can do it 💪
Shirin Khosravi Jam's engagement
Shirin Khosravi Jam's engagement rate on Instagram is 5.56%, between 3% and 6%, or 3 to 6 likes and comments per 100 followers. On a typical post that comes to about 1 like or comment for every 18 followers. Comments stay close to likes on Shirin Khosravi Jam's posts, at 1,514 likes and 701 comments on average, about 2.2 likes per comment.
Engagement rate
5.56%
- Avg likes
- 1.5K
- Avg comments
- 701
- Interactions : followers
- 1 : 18
Lookalike Creators
A free account lists creators whose audiences behave like this one, ready to shortlist and compare side by side.
Age & Gender
Unlocking shows how this audience splits across age bands and gender, charted from follower data.
Audience Location
Unlocking reveals the countries and cities where the most active followers of this account are concentrated.
Primary Language
Unlocking shows the languages this audience uses most in comments, interactions and the content it consumes.
Audience Interests
Unlocking lists the topics and categories this audience follows and engages with most often across the platform.
Authenticity
Unlocking scores how genuine the audience is, flagging fake, inactive or bot followers found in engagement patterns.
Audience Quality Score
Unlocking gives one overall audience quality score combining authenticity, engagement, follower activity and audience relevance.
Brand Affinity
Unlocking names the brands this audience already follows, mentions or engages with across social platforms.
Content Strategy
Unlocking breaks down the content pillars, tone and visual style this creator uses, and which formats perform best.
Commercial Insights
Unlocking shows collaboration types, an estimated pricing tier and a brand safety risk read for sponsored campaigns.
Industry Fit
Unlocking rates how well this creator and audience align with specific industry verticals and brand categories.
Research Shirin Khosravi Jam further
Flinque ranks Instagram creators in Technology, categories Shirin Khosravi Jam is filed under, and its free tools measure the same figures for any public account.
Related rankings
Frequently asked questions
These answers about Shirin Khosravi Jam are generated from the Instagram profile data Flinque holds for the account.
Who is Shirin Khosravi Jam?
Shirin Khosravi Jam (@jam.with.ai) is a Science, Technology & Engineering creator on Instagram, also filed under Technology. Shirin Khosravi Jam carries Instagram's verified badge and uses an Instagram professional account. The Instagram bio for Shirin Khosravi Jam runs to 20 words across 4 lines.
How many followers does Shirin Khosravi Jam have?
Shirin Khosravi Jam has 39,132 followers on Instagram (39.1K). Set against 84 accounts followed, the audience works out to about 466 followers per account Shirin Khosravi Jam follows. 89 published posts give Shirin Khosravi Jam a moderate archive on Instagram.
What is Shirin Khosravi Jam's engagement rate?
Shirin Khosravi Jam's engagement rate on Instagram is 5.56%, between 3% and 6%, or 3 to 6 likes and comments per 100 followers. On a typical post that comes to about 1 like or comment for every 18 followers. Comments stay close to likes on Shirin Khosravi Jam's posts, at 1,514 likes and 701 comments on average, about 2.2 likes per comment.
What does Shirin Khosravi Jam post about on Instagram?
Flinque files Shirin Khosravi Jam under Science, Technology & Engineering, with Technology as a secondary category. Of the 10 most recent posts shown for Shirin Khosravi Jam, 6 are videos and 4 are carousels. The 10 recent Instagram captions Flinque holds for Shirin Khosravi Jam repeatedly use the words comment, data, hope, it’s and llms. Shirin Khosravi Jam writes those captions mainly in English. Shirin Khosravi Jam keeps 1 story highlight, which open with a free Flinque account.
How do I contact Shirin Khosravi Jam for a collaboration?
Shirin Khosravi Jam's contact details are not published on Flinque's public profile. The Instagram bio links to 2 external sites, and a free Flinque account opens those links. Brands with a free Flinque account can shortlist Shirin Khosravi Jam and use Flinque's outreach tools wherever a contact route is on file.
Links
Shirin Khosravi Jam's own Instagram profile is the one outbound link Flinque publishes for the account, and the 2 bio links open after signing up.
Public profile data sourced from Instagram. Flinque is not affiliated with Shirin Khosravi Jam.