Jaydeep Shah (JD)

Senior Engineer | Edge AI & On-Device LLM Inference | Qualcomm x Google LiteRT Hackathon 2026 Winner | Building Redacto - zero-trust PII redaction with Gemma 4 on Snapdragon 8 Elite | PREQ focused

Bay Area, CA Joined on May 17, 2026

Jaydeep Shah (JD)

Jul 5

One Prompt Was Not Enough: Building a 4-Step On-Device Redaction Pipeline

#edgeai #android #litertlm #privacy

14 min read

Jaydeep Shah (JD)

Jul 5

What I Learned Getting an NPU to Actually Initialize: Six Silent Failures

#edgeai #android #litertlm #npu

13 min read

Jaydeep Shah (JD)

Jul 5

Prompt Engineering Beat My Fine-Tuned Model. Here Is Why.

#edgeai #android #litertlm

8 min read

Jaydeep Shah (JD)

Jul 4

What On-Device AI Benchmarks Actually Feel Like

#edgeai #android #litertlm

8 min read

Jaydeep Shah (JD)

Jul 4

One Model, Three Chips, Two Files: How LiteRT Delegates Really Work

#edgeai #android #litert

10 min read

Jaydeep Shah (JD)

Jun 16

What I Learned About "On-Device AI": Most of It Is a Promise, Not a Guarantee

#edgeai #android #privacy

5 min read

Jaydeep Shah (JD)

Jun 16

The Invisible Layer Between My Prompt and the Model

#edgeai #android #litertlm

4 min read

Jaydeep Shah (JD)

Jun 16

I opened a .litertlm file. Here is what is actually in there

#edgeai #android #litertlm

5 min read

Jaydeep Shah (JD)

May 18

Why My LLM Runs 4x Faster on Hardware I Had Never Heard Of

#edgeai #android #litertlm

7 min read

Jaydeep Shah (JD)

May 18

FP32, INT4, and Everything Between - What I Learned About Precision on Mobile

#deeplearning #machinelearning #mobile #performance

7 min read

Gemma 4 Challenge: Write about Gemma 4 Submission

Jaydeep Shah (JD)

May 17

What I Learned Untangling LiteRT, LiteRT-LM, and TFLite

#gemmachallenge #edgeai #android #litertlm

4 min read

Gemma 4 Challenge: Write about Gemma 4 Submission

Jaydeep Shah (JD)

May 17

What I Learned Turning a HuggingFace Model Into Something My Phone Can Run

#gemmachallenge #edgeai #android #litertlm

9 min read

Gemma 4 Challenge: Write about Gemma 4 Submission

Jaydeep Shah (JD)

May 17

What I Learned by Dissecting gemma-4-E2B-it_qualcomm_sm8750.litertlm

#litertlm #gemmachallenge #edgeai #android

8 min read

DEV Community

Jaydeep Shah (JD)

Badges

Gemma 4 Challenge Completion

Writing Debut

One Prompt Was Not Enough: Building a 4-Step On-Device Redaction Pipeline

What I Learned Getting an NPU to Actually Initialize: Six Silent Failures

Prompt Engineering Beat My Fine-Tuned Model. Here Is Why.

What On-Device AI Benchmarks Actually Feel Like

One Model, Three Chips, Two Files: How LiteRT Delegates Really Work

What I Learned About "On-Device AI": Most of It Is a Promise, Not a Guarantee

The Invisible Layer Between My Prompt and the Model

I opened a .litertlm file. Here is what is actually in there

Why My LLM Runs 4x Faster on Hardware I Had Never Heard Of

FP32, INT4, and Everything Between - What I Learned About Precision on Mobile

What I Learned Untangling LiteRT, LiteRT-LM, and TFLite

What I Learned Turning a HuggingFace Model Into Something My Phone Can Run

What I Learned by Dissecting gemma-4-E2B-it_qualcomm_sm8750.litertlm