Fission Labs (@Fission_Labs) header
Fission Labs (@Fission_Labs) profile picture

Offshore Engineering Accelerator for Technology Innovators

Sunnyvale, California, USA
Joined January 2015
Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)Photo by Fission Labs (@Fission_Labs)
โœจ๐—–๐—ฒ๐—น๐—ฒ๐—ฏ๐—ฟ๐—ฎ๐˜๐—ถ๐—ป๐—ด ๐˜๐—ต๐—ฒ ๐—™๐—ฒ๐˜€๐˜๐—ถ๐˜ƒ๐—ฎ๐—น ๐—ผ๐—ณ ๐—Ÿ๐—ถ๐—ด๐—ต๐˜๐˜€ ๐—ฎ๐˜ ๐—™๐—ถ๐˜€๐˜€๐—ถ๐—ผ๐—ป ๐—Ÿ๐—ฎ๐—ฏ๐˜€!๐Ÿช” Vibrant rangolis, traditional attire, a lot of laughter and the joy of coming together as one team. Wishing everyone a bright, joyful and prosperous Diwali! #HappyDiwali
9
An AI agent rarely fails loudly. It fails one authorized tool call at a time. Amazon Bedrock AgentCore Policy checks every tool call against a Cedar rule before it runs. The rule lives in the system, not the prompt. Breakdown: fissionlabs.com/blog-posts/aโ€ฆ #AgentCore #AgenticAI
17
Getting an AI agent to work is the easy part. Production is where teams stall. Amazon Bedrock AgentCore Harness turns much of the runtime, memory, tool and tracing work into config. Two API calls, one running agent. Our breakdown: fissionlabs.com/blog-posts/aโ€ฆ #AgentCore #AWS
29
Celebrating the festival of flowers. This week, our team celebrated Bathukamma and Dussehra. From making the Bathukamma together to fun games, traditional attire and dancing around it, the day was a celebration of tradition and togetherness. #LifeAtFissionLabs #Bathukamma
15
Fission Labs has earned the AWS Agentic AI Consulting Services Competency, bolstering our agentic AI capabilities. Thank you to @AWS_Partners, our clients and our team for making this possible. #AWS #AgenticAI #AWSPartner
22
Model capability is no longer what holds enterprise AI back. Every team can access a frontier model. What separates a working system from a demo is everything around the model See how we prioritize AI use cases: fissionlabs.com/agentic-ai-uโ€ฆ #EnterpriseAI #GenerativeAI #AIStrategy
15
We benchmarked ๐—Ÿ๐—น๐—ฎ๐—บ๐—ฎ ๐Ÿฏ.๐Ÿญ ๐Ÿด๐—• ๐—œ๐—ป๐˜€๐˜๐—ฟ๐˜‚๐—ฐ๐˜ across FP16, FP8, FP4, and NVFP4 using vLLM and llm-compressor, evaluating performance on MMLU-Pro and MATH-500. Read the full analysis fissionlabs.com/blog-posts/lโ€ฆ #LLMQuantization #LLMOps #ModelOptimization
6
๐—”๐—œ ๐—ฎ๐—ฑ๐—ผ๐—ฝ๐˜๐—ถ๐—ผ๐—ป ๐—ถ๐˜€ ๐—ฎ๐—ฐ๐—ฐ๐—ฒ๐—น๐—ฒ๐—ฟ๐—ฎ๐˜๐—ถ๐—ป๐—ด. ๐—•๐˜‚๐˜ ๐˜๐˜‚๐—ฟ๐—ป๐—ถ๐—ป๐—ด ๐—ฎ๐—ฑ๐—ผ๐—ฝ๐˜๐—ถ๐—ผ๐—ป ๐—ถ๐—ป๐˜๐—ผ ๐—ฒ๐—ป๐˜๐—ฒ๐—ฟ๐—ฝ๐—ฟ๐—ถ๐˜€๐—ฒ ๐˜ƒ๐—ฎ๐—น๐˜‚๐—ฒ ๐—ฟ๐—ฒ๐—บ๐—ฎ๐—ถ๐—ป๐˜€ ๐—ฎ ๐—ฐ๐—ต๐—ฎ๐—น๐—น๐—ฒ๐—ป๐—ด๐—ฒ. Start by identifying where your AI use cases stand: fissionlabs.com/agentic-ai-uโ€ฆ #EnterpriseAI #AgenticAI #AI
8
We are now part of the Claude Partner Network. This partnership brings together Anthropicโ€™s AI capabilities with our expertise in use-case discovery and prioritization, from pilots to production. #anthropic #anthropicpartnernetwork #enterpriseai #genai #agenticai
1
14
What happens when AI agents need access to information beyond their training data? We explored web search with Amazon Bedrock AgentCore to see how agents can retrieve real-time information and use it to generate more relevant responses. ๐Ÿ”— fissionlabs.com/blog-posts/aโ€ฆ #FissionLabs
8
9/10 security checks passed. The 1 that didnโ€™t exposed a real blind spot. AWS Security Agent flagged insufficient Log Protection: PII redaction wasnโ€™t addressed. Secure logs mean little if sensitive data gets logged. Full breakdown: fissionlabs.com/blog-posts/aโ€ฆ #AWS #CloudSecurity
11
Weโ€™re proud to be recognized by AWS for our DevOps expertise, helping businesses move faster, improve reliability, and scale with confidence. Thank you to our AWS Partner team and the Fission Labs team for their continued support and collaboration. #AWS #DevOps #FissionLabs
2
34
We're excited to announce that Fission Labs is now an AWS ๐—˜๐˜…๐—ฝ๐—ฒ๐—ฟ๐—ถ๐—ฒ๐—ป๐—ฐ๐—ฒ-๐—•๐—ฎ๐˜€๐—ฒ๐—ฑ ๐—”๐—ฐ๐—ฐ๐—ฒ๐—น๐—ฒ๐—ฟ๐—ฎ๐˜๐—ถ๐—ผ๐—ป (๐—˜๐—•๐—”) ๐—–๐—ฒ๐—ฟ๐˜๐—ถ๐—ณ๐—ถ๐—ฒ๐—ฑ ๐—ฃ๐—ฎ๐—ฟ๐˜๐—ป๐—ฒ๐—ฟ! This certification reflects our deep expertise in AWS and our commitment to driving real business outcomes. #AWSEBA
1
11
AI agent deployment on AWS just got simpler. ๐Ÿš€ With Amazon Bedrock AgentCoreโ€™s CodeZip, skip Dockerfiles, ARM builds & ECR. We deployed a production multi-agent pipeline with updates taking ~10 seconds. Read more: fissionlabs.com/blog-posts/aโ€ฆ #Fissionlabs #AWS #AgentCore
8
AI agents are reshaping web traffic. AWS WAFโ€™s AI Traffic Analysis Dashboard helps organizations understand AI agent activity and make smarter decisions around security, infrastructure costs, and content monetization. Read more: fissionlabs.com/blog-posts/mโ€ฆ #Fissionlabs #AIAgents
7
Security testing is shifting from a final-stage checkpoint to a continuous part of the software development lifecycle. In our latest blog, we look at AWS Security Agent, what impressed us, where we remain cautious, and who should consider trying it. fissionlabs.com/blog-posts/aโ€ฆ #AWS
1
30
Weโ€™re excited to share that Fission Labs has been named an OpenAI Select Partner. Weโ€™re proud to be part of this next chapter. Learn more: openai.com/business/partnersโ€ฆ #OpenAISelectPartner #OpenAIPartnerNetwork #EnterpriseAI
2
18
Today, we celebrate the spirit of freedom, resilience, and the countless individuals whose vision shaped the India we proudly call home. ๐Ÿ‡ฎ๐Ÿ‡ณ Happy Independence Day! #FissionLabs #HappyIndependenceDay #India #Innovation #Technology #DigitalTransformation #AI #NationBuilding
1
10
A respectful workplace is built through awareness, empathy, and shared responsibility. At Fission Labs, we recently conducted a POSH Awareness Session to reinforce our commitment to fostering a safe, inclusive, and respectful work environment for everyone. #POSH #RespectAtWork
14
Gartner projects 60% of AI projects will be abandoned through 2026. The model usually gets blamed. The real cause is six production dimensions nobody checks before the budget is committed: accuracy, safety, evaluation, access control, scale, tokenomics. fissionlabs.com/blog-posts/tโ€ฆ
12