Opportunity Detail

Questions and Answers

UC IVR Speech to Text Technology Project
Document #:  JFSRLBOIS12025

Question:   1. In the RLB, ODJFS references "ODJFS has development work underway to collect work search information by recording voice response on the IVR". Does the development include existing IVR UC Technology? If so, what platform? 2. How many simultaneous IVR sessions does this solution need to support? 3. How many records / Individuals are in the search Database? 4. Does this solution need to tie into an existing phone solution. If so, what platform? 5. Are example resumes acceptable, or does ODJFS require specific individuals who will be able to interview and perform if awarded? 6. How many resumes are required to be included in the proposal?

Answer:   1. On page 6, Section VI, Subsection A - General Scope: "The State of Ohio UC IVR is currently built on a Cisco CVP 8.0 platform. The claimant records are stored on a local SQL 2008 clustered environment which generates a flat file as part of a nightly batch that is collected by the State’s host system for adjudication. 2. For this project the state does not expect direct integration with the IVR sessions. The "General Scope" section on pages 6 and 7 contains information to be used for designing and sizing the solution. Information to be used specifically for sizing includes: • Peak hourly recordings are estimated at 10,000. • Estimated yearly volumes are estimated to be 4.5 million recordings. • Recorded file sizes range from 350k to 450k in size, lasting 30-45 seconds. • All recordings must complete transcription within 30 minutes of receipt of recording. 3. There are 2,782,189 claimant records. 4. No. 5. Specific individuals must be included for key staff as described on page 5. “The vendor must demonstrate significant expertise by assigning staff to key leadership roles for this project. The vendor must, at minimum identify, by position and by name, those staff they consider key to the projects success, and professional profiles and resumes/curriculum vitae must be included in vendor proposals for all persons proposed for key positions. At minimum, key staff identified must include a project manager assigned to the project for the entire term of the project who will work with the ODJFS overall project manager, and a technical lead with at least three years speech to text technology design experience.” 6. Again, as described on page 5. "At minimum, key staff identified must include a project manager assigned to the project for the entire term of the project who will work with the ODJFS overall project manager, and a technical lead with at least three years speech to text technology design experience."

Date: 2/7/2012

Inquiry: 24852

Question:   can you provide a specific example of a recorded audio file? we are trying to understand what exactly is included with the series of recorded questions. Does each file include audio file of every question and the answer?

Answer:   Two sample files have been provided. To obtain these sample audio files, please send your request to the following email address: OIS-IT-BID-REQUEST@jfs.ohio.gov . Please enter the following in the subject line: Request for Wave Files_JFSRLBOIS12025.

Date: 2/7/2012

Inquiry: 24851

Question:   -Can DJFS provide the street address from the State Directory Assistance (i.e. 411) (before the bid being awarded)? -Can you provide a list of businesses within the State (before the bid being awarded)? -How many audio files are created per call? -Would you be amenable to reconfigure your IVR., considering front end speech recognition solution for IVR? -What percent of traffic is Spanish? -Can we get the actual questions asked the callers from the IVR (to help us understand if there are categories or free form audio)? -Do the callers provide a pin number (maybe social security number) that can identify who the caller is?

Answer:   Questions: a. Can DJFS provide the street address from the State Directory Assistance (i.e. 411) (before the bid being awarded)? b. Can you provide a list of businesses within the State (before the bid being awarded)? c. How many audio files are created per call? d. Would you be amenable to reconfigure your IVR., considering front end speech recognition solution for IVR? e. What percent of traffic is Spanish? f. Can we get the actual questions asked the callers from the IVR (to help us understand if there are categories or free form audio)? g. Do the callers provide a pin number (maybe social security number) that can identify who the caller is? Answers: a. We cannot provide this information as requested. b. We cannot provide this information as requested. c. One d. Please ensure your bid meets the requirements specified in the RLB. e. The percentage of Spanish recordings for the Month of January, 2012 is less than 1%. f. This will be provided during the demo phase of the process. g. Yes.

Date: 2/3/2012

Inquiry: 24847

Question:   In section VII, Format of Submission (pg 9), can you clarify how many letterhead bids must be submitted? Is it 3 or 5?

Answer:   Five (5) vendor proposals are to be submitted (1-original and 4-copies)

Date: 2/3/2012

Inquiry: 24846

Question:   Under Section VII, Format of Submission, can you clarify how many copies must be submitted? Is it 3 or 5?

Answer:   Please submit five (5) vendor proposals (1 - original and 4 - copies).

Date: 2/3/2012

Inquiry: 24845

Question:   When will previous questions be answered? My technical team needs to make architecture decisions and are waiting on answers to previous questions.

Answer:   Today, February 3, 2012

Date: 2/3/2012

Inquiry: 24844

Question:   Page 4 #3. parse one wave file into multiple elements. Question: Does your system have the ability to "not" record one continuous file, but to record each question, then post just that recording so that there are multiple audio files that we are using instead of one. Question 2: Does your system have the ability to insert a delimiter after each question such as a long tone, or a spoken word as a marker that allows for us to recognize where a question starts and stops. Thank you

Answer:   1.) At the present time the system provides one file. The proposal should be made with the assumption this will not change. 2.)Each file is recorded with a series standard pre-recorded phrase question, followed by captured recording.

Date: 2/3/2012

Inquiry: 24833

Question:   Where do we include pricing for the solution? Do we include as part of Analysis and Design? Do you only need a total cost figure or are we to list out the solution elements and detail somewhere in our response?

Answer:   In Tab D of the submission (Vendor Scope response) the vendor should include the detailed specifications (quantity, part number, description, STS price) of recommended hardware and software to be purchased for the solution as discussed on page 6. The submitted recommendations should include a current STS Number from which the hardware and software components can be purchased. The State will evaluate the recommended hardware and software against available existing infrastructure and work with the vendor during the design phase to finalize the hardware and software specifications. A subsequent competitive procurement will be undertaken by the State to acquire the hardware and additional software based on the agreed-to final design. A revised cost sheet has been provided in Addendum A, which will replace the Cost Summary Sheet on page 37. The vendor should use the revised cost sheet for submitting all proposed costs, which should include analysis and design, installation, testing, training, implementation, Speech-to-Text software and one year annual maintenance for STT software.

Date: 2/3/2012

Inquiry: 24832

back

Inquiry period ended:  2/6/2012 8:00:00 AM